JP2022500965A - Video signal coding / decoding method and equipment for that purpose - Google Patents

Abstract

The embodiments of the present invention provide video signal processing methods and devices. In particular, a method of decoding a video signal based on a reduced transform according to an embodiment of the present invention points to transform kernels that are applied horizontally and vertically to the current block. The area to which the transformation is applied to the current block is determined based on the stage at which the transform index is obtained from the video signal and the size of the transform kernel and the current block indicated by the transform index. The stage in which the transformation is applied, the stage in which the coefficient of the remaining region other than the region to which the transformation is applied is regarded as 0, and the transformation instructed by the transformation index for the region to which the transformation is applied. The kernel can be used to include the step of performing an inverse transform. [Selection diagram] Fig. 1

Description

The embodiments of the present specification relate to a method and an apparatus for processing a video signal, and further, for encoding / decoding a video signal by performing a conversion based on a specifically reduced conversion. Regarding methods and equipment.

Compression coding (encoding) means a series of signal processing techniques for transferring digitized information via a communication line and storing it in a form suitable for a storage medium. Media such as video, video, and audio can be the target of compression coding, and in particular, a technique for performing compression coding on video is called video video compression.

Next-generation video content will be characterized by high spatial resolution, high frame rate, and high dimensionality of scene representation. Processing such content results in enormous increases in memory storage, memory access rate and processing power.

Therefore, it is necessary to design a coding tool for processing next-generation video content more efficiently. In particular, the video codec standard after the HEVC (high efficiency video coding) standard is efficient for converting video signals in the spatial domain into the frequency domain, along with predictive technology with even higher accuracy. Requires a high conversion technology.

An object of the embodiments of the present specification is to propose a method of performing a primary transform on a predefined region according to a specific condition.

The technical problems to be solved in the examples of the present specification are not limited to the technical problems mentioned above, and other technical problems not mentioned above are described below in the technical field to which the present invention belongs. It will be clearly understandable to those with normal knowledge.

The uniform phase of the examples herein is the transform kernels applied in the horizontal and vertical directions of the current block in the method of decoding video signals based on reduced transforms. The area to which the transformation is applied to the current block based on the step of acquiring the indicated transform index from the video signal and the size of the transform kernel and the current block indicated by the transform index. Is indicated by the transformation index for the step of determining, the step of considering the coefficient of the remaining region other than the region to which the transformation is applied as 0 in the current block, and the region to which the transformation is applied. The transformation kernel can be used to include the steps of performing an inverse transform.

Preferably, the step of determining the region to which the transformation applies is a transformation in which the transformation kernel indicated by the transformation index is predefined, the width and / or height of the current block. If height) is greater than the predefined size, this can be done by determining a region having the width and / or height of the predefined size as the region to which the transformation applies.

Preferably, the predefined conversion can be any one of a plurality of conversion combinations configured with a combination of DST7 and / or DCT8.

Preferably, the predefined size can be 16.

Preferably, the step of determining the region to which the transformation applies is the smaller of the current block widths and 32 if the transformation kernel indicated by the transformation index belongs to the first transformation group. Is determined to be the width of the region to which the transformation is applied, and the smaller of the current block height (height) and 32 is determined to be the height of the region to which the transformation is applied, and the transformation is performed. If the conversion kernel indicated by the index belongs to the second conversion group, the smaller value of the current block width and 16 is determined to be the width of the area to which the conversion is applied, and the height of the current block. This can be done by determining a smaller value of 16 to the height of the region to which the transformation is applied.

Preferably, the coefficient to which the inverse transformation is applied further comprises the step of acquiring a syntax element indicating the position of the last signifi-cant coefficient in the current block. Can be obtained from the video signal based on the position of the effective factor of.

Preferably, the syntax element is binarized in a truncated unary manner, and the maximum value of the syntax element can be determined based on the region considered to be zero.

A uniform phase of an embodiment of the present specification includes a memory for storing the video signal and a processor coupled to the memory in a device for decoding a video signal based on a reduced transform. , The processor obtains a transform index from the video signal indicating the transform kernels applied in the horizontal and vertical directions of the current block, and the transform kernel and the transform kernel indicated by the transform index. Based on the size of the current block, the region to which the transformation is applied to the current block is determined, and the coefficient of the remaining regions other than the region to which the transformation is applied in the current block is set to 0. It can be considered and an inverse transform can be performed on the region to which the transformation is applied using the transformation kernel indicated by the transformation index.

According to the embodiments of the present specification, the complexity can be significantly reduced by performing the conversion only on the predefined regions according to specific conditions.

The effects obtained in the examples of the present specification are not limited to the effects mentioned above, and other effects not mentioned above are clearly described by the following description to those who have ordinary knowledge in the technical field to which the present invention belongs. You can understand.

The accompanying drawings, included as part of a detailed description to aid in understanding of the invention, provide embodiments of the invention and illustrate the technical features of the invention along with the detailed description.

An example of a video coding system is shown as an embodiment to which the present invention is applied. As an embodiment to which the present invention is applied, a schematic block diagram of an encoding device in which a video / video signal is encoded is shown. As an embodiment to which the present invention is applied, a schematic block diagram of a decoding device in which a video signal is decoded is shown. As an embodiment to which the present invention is applied, it is a structural diagram of a content streaming system. As embodiments to which the present invention can be applied, FIG. 5a is a QT (Quad Tree, hereinafter referred to as “QT”), FIG. 5b is a BT (Binary Tree, hereinafter referred to as “BT”), and FIG. TT (Ternary Tree, hereinafter referred to as “TT”) FIG. 5d is a diagram for explaining a block division structure by AT (Asymmetric Tree, hereinafter referred to as “AT”). As an embodiment to which the present invention is applied, a schematic block diagram of a conversion and quantization unit and an inverse quantization and inverse conversion unit in an encoding device is shown. As an embodiment to which the present invention is applied, FIG. 7 shows a schematic block diagram of an inverse quantization and inverse transformation unit in a decoding apparatus. It is a flowchart which shows the process which AMT (adaptive multiple transform) is executed. It is a flowchart which shows the decoding process which AMT is executed. It is a flowchart which shows the inverse transformation process based on MTS based on the embodiment of this invention. It is a block diagram of the apparatus which performs the decoding based on MTS based on the embodiment of this invention. As an embodiment to which the present invention is applied, it is an encoding / decoding flowchart to which a secondary transformation is applied. As an embodiment to which the present invention is applied, it is an encoding / decoding flowchart to which a secondary transformation is applied. As an embodiment to which the present invention applies, FIG. 14 shows a diagram for explaining Givens rotation. As an embodiment to which the present invention is applied, a one-round configuration with a 4x4 NSST (non-separable secondary transform) composed of a given rotation layer and permutation is shown. As an embodiment to which the present invention is applied, the operation of RST (reduced secondary transform) is shown. As an embodiment to which the present invention is applied, it is a figure which shows the process of performing the reverse scan from the 64th to the 17th based on the reverse scan order. As an embodiment to which the present invention is applied, an example of an encoding flowchart using a single transform indica-tor (STI) is shown. As an embodiment to which the present invention is applied, an example of an encoding flowchart using a unified transform in-dicator (UTI) is shown. As an embodiment to which the present invention is applied, another example of an encoding flowchart using UTI is shown. As an embodiment to which the present invention is applied, another example of an encoding flowchart using UTI is shown. As an embodiment to which the present invention is applied, an example of an encoding flowchart for executing conversion is shown. As an embodiment to which the present invention is applied, an example of a decoding flowchart for executing conversion is shown. As an embodiment to which the present invention is applied, an example of a detailed block diagram of a conversion unit 120 in the encoding device 100 is shown. As an embodiment to which the present invention is applied, an example of a detailed block diagram of the inverse conversion unit 230 in the decoding device 200 is shown. As an embodiment to which the present invention is applied, a flowchart for processing a video signal is shown. It is a flowchart which illustrates the conversion method of the video signal by embodiment to which this invention is applied. As an example to which the present invention is applied, it is a figure illustrating a method of encoding a video signal by using a reduced transform. As an example to which the present invention is applied, it is a figure illustrating a method of decoding a video signal by using a reduced transform. It is a flowchart which illustrates the method of decoding the video signal based on the reduced transform which concerns on embodiment of this specification. As an example to which the present invention is applied, an example of a block diagram of an apparatus for processing a video signal is shown.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. The detailed description disclosed below, along with the accompanying drawings, is intended to illustrate exemplary embodiments of the invention and is intended to show only embodiments in which the invention can be practiced. Not. The following detailed description includes specific details to provide a complete understanding of the invention. However, those skilled in the art will appreciate that the present invention can be carried out without such specific details.

In some cases, known structures and devices may be omitted or shown in the form of block diagrams centered on the core functions of each structure and device, in order to avoid obscuring the concepts of the invention. can.

Further, as the term used in the present invention, a general term widely used at present is selected as much as possible, but in a specific case, a term arbitrarily selected by the applicant will be used for explanation. In such a case, in order to clearly state the meaning in the detailed description of the relevant part, the term should not be simply interpreted only by the name of the term used in the description of the present invention. It should be clarified that the meaning of is to be understood and interpreted.

The specific terms used in the following description are provided to aid in the understanding of the invention, and the use of such specific terms is in other forms to the extent that it does not deviate from the technical ideas of the invention. Can be changed to. For example, signals, data, samples, pictures, frames, blocks, etc. may be appropriately substituted and interpreted in each coding process.

Hereinafter, "processing unit" as used herein means a unit on which an encoding / decoding processing process such as prediction, conversion, and / or quantization is performed. Further, the processing unit can be interpreted as including a unit of a luminance (luma) component and a unit of a color difference (chroma) component. For example, the processing unit can be a block, a coding unit (CU), a prediction unit (PU), or a transform unit (TU).

Further, the processing unit can be interpreted as a unit of a luminance component or a unit of a color difference component. For example, the processing unit can correspond to a coding tree block (CTB), a coding block (coding block, CB), a PU or a transform block (TB) of a luminance component. Alternatively, the processing unit can correspond to the color difference components CTB, CB, PU, and TB. Further, the processing unit is not limited to this, and may be interpreted as including a unit of a luminance component and a unit of a color difference component.

Further, the processing unit is not necessarily limited to a square block, and may be configured in the form of a polygon having three or more vertices.

Hereinafter, in the present specification, pixels or pixels are commonly referred to as samples. Then, using a sample can mean using a pixel value, a pixel value, or the like.

FIG. 1 shows an example of a video coding system as an embodiment to which the present invention is applied.

The video coding system can include a source device 10 and a receiving device 20. The source device 10 can transfer the encoded video / video information or data to the receiving device 20 via a digital storage medium or network in the form of a file or streaming.

The source device 10 can include a video source 11, an encoding device 12, and a transmitter 13. The receiving device 20 can include a receiver 21, a decoding device 22, and a renderer 23. The encoding device 10 can be called a video / video encoding device, and the decoding device 20 can be called a video / video decoding device. The transmitter 13 can be included in the encoding device 12. The receiver 21 can be included in the decoding device 22. The renderer 23 may include a display unit and may be composed of another device or an external component of the display unit.

Video sources can acquire video / video through video / video capture, compositing or generation processes, etc. Video sources can include video / video capture devices and / or video / video generation devices. The video / video capture device can include, for example, one or more cameras, a video / video archive containing previously captured video / video, and the like. Video / video generation devices can include, for example, computers, tablets, smartphones, etc., and can (electronically) generate video / video. For example, a virtual video / video can be generated, such as through a computer, in which case the video / video capture process can be replaced in the process of generating the relevant data.

The encoding device 12 can encode the input video / video. The encoding device 12 can perform a series of procedures such as prediction, conversion, and quantization for compression and coding efficiency. The encoded data (encoded video / video information) can be output in the form of a bitstream.

The transfer unit 13 can transfer the encoded video / video information or data output in the form of a bitstream to the receiving unit of the receiving device via a digital storage medium or network in the form of a file or streaming. The digital storage medium can include various storage media such as USB, SD, CD, DVD, Blu-ray, HDD, SSD and the like. The transfer unit 13 can include an element for generating a media file via a predetermined file format, and can include an element for transfer via a broadcast / communication network. The receiver 21 can extract the bitstream and reach the decoding device 22.

The decoding device 22 can perform a series of procedures such as inverse quantization, inverse transformation, and prediction corresponding to the operation of the encoding device 12 to decode the video / video.

The renderer 23 can render the decoded video / video. The rendered video / video can be displayed via the display unit.

FIG. 2 shows a schematic block diagram of an encoding device in which video / video signal encoding is performed, as an embodiment to which the present invention is applied. The encoding device 100 of FIG. 2 can correspond to the encoding device 12 of FIG.

The video dividing unit 110 can divide the input image (or picture, frame) input to the encoding device 100 into one or more processing units. As an example, the processing unit can be referred to as a coding unit (CU). In this case, the coding unit is recursively divided from the coding tree unit (CTU) or the largest coding unit (LCU) based on the QTBT (Quad-tree binary-tree) structure. Can be done. For example, a coding unit can be divided into multiple coding units of deeper depth based on a quad tree structure and / or a binary tree structure. In this case, for example, the quad tree structure can be applied first and the binary tree structure can be applied later. Alternatively, the binary tree structure may be applied first. The coding procedure according to the present invention can be executed based on the final coding unit which is not further divided. In this case, the largest coding unit can be immediately used as the final coding unit, based on coding efficiency etc. according to the characteristics of the video, or if necessary, the coding unit is recursively lower than recursively. It is divided into depth coding units, and the optimum size coding unit is used as the final coding unit. Here, the coding procedure can include procedures such as prediction, conversion, and restoration described later. As another example, the processing unit may further include a prediction unit (PU: Prediction Unit) or a transformation unit (TU: Transform Unit). In this case, the prediction unit and the conversion unit can be partitioned or partitioned from the final coding unit described above, respectively. The prediction unit may be a unit of sample prediction, and the conversion unit may be a unit for deriving a conversion coefficient and / or a unit for deriving a residual signal from the conversion coefficient.

Units are sometimes used in combination with terms such as block or area. In the general case, an MxN block can represent a sample or set of transform coefficients consisting of M columns and N rows. The sample may generally show a pixel or pixel value, may only show the pixel / pixel value of the luminance (luma) component, or only the pixel / pixel value of the saturation (chroma) component. You can also. In the sample, one picture (or video) is used as a term corresponding to a pixel or a pel.

The encoding device 100 subtracts the prediction signal (predicted block, prediction sample array) output from the inter-prediction unit 180 or the intra-prediction unit 185 from the input video signal (original block, original sample array) to obtain a resilient signal (predictive block, prediction sample array). (Remaining signal, remaining blocks, remaining sample array) can be generated, and the generated residency signal is transferred to the conversion unit 120. In this case, as shown in the figure, the unit that subtracts the predicted signal (predicted block, predicted sample array) from the input video signal (original block, original sample array) in the encoder 100 can be called the subtraction unit 115. The prediction unit can predict a block to be processed (hereinafter referred to as a current block) and generate a predicted block (predicted block) including a prediction sample of the current block. The predictor can currently determine whether intra-forecasts are applied or inter-forecasts are applied on a block or CU basis. As will be described later in the description of each prediction mode, the prediction unit can generate various information related to prediction such as prediction mode information, and can reach the entropy encoding unit 190. Information about prediction can be encoded by the entropy encoding unit 190 and output in the form of a bitstream.

The intra prediction unit 185 can predict the current block by referring to the sample in the current picture. The referenced sample can be located near or far from the current block, depending on the prediction mode. In intra-prediction, the prediction mode can include a plurality of non-directional modes and a plurality of directional modes. The non-directional mode can include, for example, a DC mode and a planner mode (Planar mode). The directional mode can include, for example, 33 directional prediction modes or 65 directional prediction modes, depending on the degree of fineness of the prediction direction. However, as an example, more or less directional prediction modes can be used depending on the setting. The intra prediction unit 185 can also determine the prediction mode currently applied to the block by using the prediction mode applied to the peripheral block.

The inter-prediction unit 180 can derive the predicted block of the current block based on the reference block (reference sample array) identified by the motion vector on the reference picture. At this time, in order to reduce the amount of motion information transferred in the inter-prediction mode, the motion information is predicted in block, sub-block, or sample units based on the correlation of the motion information between the surrounding block and the current block. Can be done. The motion information can include a motion vector and a reference picture index. The motion information can further include information in the inter-prediction direction (L0 prediction, L1 prediction, Bi prediction, etc.). For inter-prediction, the peripheral block can include a spatial neighboring block present in the current picture and a temporal neighboring block present in the reference picture. A reference picture containing a reference block and a reference picture containing a temporal peripheral block may be the same or different. The temporal peripheral block can be called by the same name such as a collocated reference block or the same position (CU colCU), and the reference picture including the temporal peripheral block can be referred to as a collocated picture, colPic. ). For example, the inter-prediction unit 180 constitutes a motion information candidate list based on the peripheral block, and generates information indicating which candidate is used to derive the motion vector and / or the reference picture index of the current block. can do. Inter-prediction can be performed based on various prediction modes. For example, in the case of skip mode and merge mode, the inter-prediction unit 180 can use the motion information of the peripheral block as the motion information of the current block. .. In skip mode, unlike merge mode, the residency signal may not be transferred. In the motion vector prediction (MVP) mode, the motion vector of the peripheral block is used by the motion vector predictor to signal the motion vector difference of the current block. You can specify the motion vector.

The prediction signal generated via the inter-prediction unit 180 or the intra-prediction unit 185 is used to generate a restoration signal or is used to generate a residency signal.

The transform unit 120 can apply a transform method to the resilient signal to generate transform coefficients. For example, the transformation technique is at least one of DCT (Discrete Cosine Transform), DST (Discrete Sine Transform), KLT (Karhunen-Loeve Transform), GBT (Graph-Based Transform), or CNT (Conditionally Non-linear Transform). Can include one. Here, GBT means a transformation obtained from this graph when the relationship information between pixels is represented by a graph. CNT means a transformation obtained based on a predictive signal generated using all previously reconstructed pixels. The conversion process may also be applied to pixel blocks of the same size as a square, and may also be applied to non-square variable size blocks.

The quantized unit 130 quantizes the conversion coefficient and transfers it to the entropy encoding unit 190, and the entropy encoding unit 190 encodes the quantized signal (information about the quantized conversion coefficient) and outputs it as a bit stream. can do. Information about the quantized conversion coefficients can be referred to as resilient information. The quantization unit 130 can rearrange the quantized conversion coefficients in the form of blocks in the form of a one-dimensional vector based on the coefficient scan order, and is quantized in the form of a one-dimensional vector. It is also possible to generate information about the quantized conversion coefficient based on the conversion coefficient. The entropy encoding unit 190 can execute various encoding methods such as exponential Golomb, CAVLC (context-adaptive variable length coding), and CABAC (context-adaptive binary arithmetic coding). The entropy encoding unit 190 can encode the information necessary for video / video restoration (for example, the value of syntax elements) together or separately, in addition to the quantized conversion coefficient. Encoded information (eg, video / video information) can be transferred or stored in NAL (network abstraction layer) units in the form of bitstreams. The bitstream can be transferred over a network or stored in a digital storage medium. Here, the network can include a broadcasting network and / or a network, and the digital storage medium may include various storage media such as USB, SD, CD, DVD, Blu-ray, HDD, SSD, and the like. can. In the signal output from the entropy encoding unit 190, the transmission unit (not shown) for transmitting and / or the storage unit (not shown) for storing can be configured as an inner / outer element of the encoding device 100. , Or the transfer unit may be a component of the entropy encoding unit 190.

The quantized conversion coefficient output from the quantized unit 130 can be used to generate a prediction signal. For example, the quantized conversion coefficient can restore the resilient signal by applying the inverse quantization and the inverse transformation via the inverse quantization unit 140 and the inverse transformation unit 150 in the loop. The addition unit 155 adds a restored residency signal to the prediction signal output from the inter-prediction unit 180 or the intra-prediction unit 185 to generate a reconstructed signal (reconstructed picture, restored block, restored sample array). Can be done. As in the case where the skip mode is applied, if there is no register dual of the block to be processed, the predicted block is used as the restore block. The addition unit 155 can be called a restoration unit or a restoration block generation unit. The generated restoration signal is used for intra-prediction of the next block to be processed in the current picture, and may be used for inter-prediction of the next picture after filtering as described later.

The filtering unit 160 can apply filtering to the restored signal to improve the subjective / objective image quality. For example, the filtering unit 160 can apply various filtering methods to the restored picture to generate a modified (modified) restored picture, and can transfer the modified (restored picture to the decrypted picture buffer 170). Various filtering methods can include, for example, diblocking filtering, sample adaptive offset, adaptive loop filter, bilateral filter, and the like. Filtering unit 160. Can generate various information about filtering and reach the entropy encoding unit 190 as described later in the description of each filtering method. The information about filtering is encoded by the entropy encoding unit 190 and output in the form of a bit stream. can do.

The modified restored picture transferred to the decoded picture buffer 170 is used as a reference picture by the inter-prediction unit 180. As a result, the encoding device can avoid a prediction mismatch between the encoding device 100 and the decoding device when inter-prediction is applied, and can also improve the encoding efficiency.

The decoded picture buffer 170 can store the modified restored picture for use as a reference picture from the interprediction unit 180.

FIG. 3 shows a schematic block diagram of a decoding device in which a video signal is decoded as an embodiment to which the present invention is applied. The decoding device 200 of FIG. 3 can correspond to the decoding device 22 of FIG.

Referring to FIG. 3, the decoding device 200 includes an entropy decoding unit 210, an inverse quantization unit 220, an inverse conversion unit 230, an addition unit 235, a filtering unit 240, a decoding picture buffer (DPB) 250, an inter prediction unit 260, and an intra prediction unit. It can be composed of including a part 265. The inter-prediction unit 260 and the intra-prediction unit 265 can be collectively referred to as a prediction unit. That is, the prediction unit can include an inter prediction unit 180 and an intra prediction unit 185. The inverse quantization unit 220 and the inverse conversion unit 230 can be collectively referred to as a resilient processing unit. That is, the resilient processing unit can include an inverse quantization unit 220 and an inverse conversion unit 230. The entropy decoding unit 210, the inverse quantization unit 220, the inverse conversion unit 230, the addition unit 235, the filtering unit 240, the inter-prediction unit 260, and the intra-prediction unit 265 have one hardware component (for example, for example, depending on the embodiment). It can be configured by a decoder or processor). Also, the decoding picture buffer 250 can be implemented by one hardware component (eg, memory or digital storage medium), depending on the embodiment.

When a bitstream containing video / video information is input, the decoding device 200 can restore the video corresponding to the process in which the video / video information from the encoding device 100 of FIG. 2 is processed. For example, the decoding device 200 can perform decoding using the processing unit applied by the encoding device 100. Thus, the decoding processing unit can be, for example, a coding unit, which can be separated from the coding tree unit or the largest coding unit along a quad tree structure and / or a binary tree structure. Then, the restored video signal decoded and output via the decoding device 200 can be reproduced through the playback device.

The decoding device 200 can receive the signal output from the encoding device 100 of FIG. 2 in the form of a bitstream, and the received signal can be decoded via the entropy decoding unit 210. For example, the entropy decoding unit 210 can perform farthing (analysis) of a bitstream to derive information (for example, video / video information) necessary for video restoration (or picture restoration). For example, the entropy decoding unit 210 decodes the information in the bitstream based on a coding method such as exponential Golomb coding, CAVLC or CABAC, and the values of the syntax elements required for video restoration and the quantum of the conversion coefficient related to the residency. The converted value can be output. More specifically, the CABAC entropy decoding method receives the corresponding bin for each syntax element from the bitstream, and the information of the syntax element to be decoded and the decoding information of the periphery and the block to be decoded, or the symbol / symbol decoded in the previous step. The information in the bins is used to determine the context model, the probability of bin occurrence is predicted based on the determined context model, and the bin's arithmetic decoding is performed for each syntax. The corresponding symbol can be generated for the value of the element. At this time, the CABAC entropy decoding method can update the context model by using the symbol / bin information decoded for the next symbol / bin context model after determining the context model. Among the information decoded by the entropy decoding unit 210, the information related to the prediction is provided by the prediction unit (inter prediction unit 260 and the intra prediction unit 265), and the entropy decoding unit 210 has performed the entropy decoding, that is, the coefficient dual value. The quantized conversion coefficient and related parameter information can be input to the inverse quantization unit 220. Further, among the information decoded by the entropy decoding unit 210, the information related to filtering can be provided to the filtering unit 240. On the other hand, a receiving unit (not shown) that receives the signal output from the encoding device 100 may be further configured as an internal / external element of the decoding device 200, or the receiving unit is a component of the entropy decoding unit 210. It is possible.

The dequantization unit 220 can dequantize the quantized conversion coefficient and output the conversion coefficient. The dequantization unit 220 can rearrange the quantized conversion coefficients into the form of a two-dimensional block. In this case, the rearrangement can be performed based on the coefficient scan order performed by the encoding device 100. The inverse quantization unit 220 can perform inverse quantization of the quantized transformation coefficient using a quantization parameter (for example, quantization step size information) and acquire a transformation coefficient.

The inverse conversion unit 230 reversely transforms the conversion coefficient to acquire a register dual signal (residual block, register dual sample array).

The prediction unit can predict the current block and generate a predicted block including a prediction sample of the current block. Based on the information about the prediction output from the entropy decoding unit 210, the prediction unit can determine whether the intra prediction is applied to the current block or the inter prediction is applied, and the specific intra prediction can be applied. / Inter prediction mode can be determined.

The intra prediction unit 265 can predict the current block by referring to the sample in the current picture. The referenced sample can be located near or far from the current block, depending on the prediction mode. In intra-prediction, the prediction mode can include a plurality of non-directional modes and a plurality of directional modes. The intra prediction unit 265 can also determine the prediction mode currently applied to the block by using the prediction mode applied to the peripheral block.

The inter-prediction unit 260 can derive the predicted block of the current block based on the reference block (reference sample array) identified by the motion vector on the reference picture. At this time, in order to reduce the amount of motion information transferred in the inter-prediction mode, the motion information is predicted in block, sub-block, or sample units based on the correlation of the motion information between the surrounding block and the current block. Can be done. The motion information can include a motion vector and a reference picture index. The motion information can further include information in the inter-prediction direction (L0 prediction, L1 prediction, Bi prediction, etc.). For inter-prediction, the peripheral block can include a spatial neighboring block present in the current picture and a temporal neighboring block present in the reference picture. For example, the inter-prediction unit 260 can configure a motion information candidate list based on peripheral blocks and derive a motion vector and / or a reference picture index of the current block based on the received candidate selection information. Inter-prediction can be made based on various prediction modes, and the information about the prediction can include information indicating the mode of inter-prediction of the current block.

The addition unit 235 adds the acquired residency signal to the prediction signal (predicted block, prediction sample array) output from the inter-prediction unit 260 or the intra-prediction unit 265, whereby the restoration signal (reconstruction picture, restoration block, Restoration, sample array) can be generated. As in the case where the skip mode is applied, if there is no register dual of the block to be processed, the predicted block is used as the restore block.

The addition unit 235 can be called a restoration unit or a restoration block generation unit. The generated restore signal is used for intra-prediction of the next block to be processed in the current picture, and can also be used for inter-prediction of the next picture after filtering as described later.

The filtering unit 240 can improve the subjective / objective image quality by applying the filtering to the restored signal. For example, the filtering unit 240 can apply various filtering methods to the restored picture to generate a modified restored picture, and can transfer the modified restored picture to the decoding picture buffer 250. Various filtering methods can include, for example, diblocking filtering, sample adaptive offset (SAO), adaptive loop filter (ALF), bilateral filter and the like.

The modified restored picture transferred to the decoded picture buffer 250 is used as a reference picture by the inter-prediction unit 260.

In the present specification, the embodiments described by the filtering unit 160, the inter prediction unit 180, and the intra prediction unit 185 of the encoding device 100 are also described in the filtering unit 240, the inter prediction unit 260, and the intra prediction unit 265 of the respective decoding devices. Can be applied identically or correspondingly.

FIG. 4 is a structural diagram of a content streaming system as an embodiment to which the present invention is applied.

The content streaming system to which the present invention is applied can largely include an encoding server 410, a streaming server 420, a Web server 430, a media storage 440, a user device 450, and a multimedia input device 460.

The encoding server 410 is responsible for compressing content input from multimedia input devices such as smartphones, cameras, and camcoders into digital data to generate a bitstream, which is then transferred to the streaming server 420. As another example, if the multimedia input device 460, such as a smartphone, camera, camcorder, etc., directly generates a bitstream, the encode server 410 can be omitted.

The bitstream can be generated by the encoding method to which the present invention is applied or the method of generating the bitstream, and the streaming server 420 temporarily stores the bitstream in the process of transferring or receiving the bitstream. Can be done.

The streaming server 420 transfers multimedia data to the user apparatus 450 based on the request of the user via the Web server 430, and the Web server 430 acts as an intermediary for informing the user of what kind of service is available. .. When the user requests the desired service from the web server 430, the web server 430 transmits this to the streaming server 420, and the streaming server 420 transfers the multimedia data to the user. At this time, the content streaming system may include another control server, in which case the control server serves to control commands / responses between each device in the content streaming system.

The streaming server 420 can receive content from the media storage 440 and / or the encoding server 410. For example, when the content is to be received from the encode server 410, the content can be received in real time. In this case, in order to provide a smooth streaming service, the streaming server 420 can store the bitstream for a certain period of time.

Examples of the user device 450 include mobile phones, smartphones (smart phones), laptop computers, digital broadcasting terminals, PDA (personal digital assistants), PMPs (portable multimedia players), navigation systems, and slate PCs (slate). PC), tablet PC (tablet PC), ultrabook, wearable device (eg, watch-type terminal (smartwatch), glass-type terminal (smart glass), HMD (head mounted display)), digital TV, You can have a desktop computer, digital signage, etc.

Each server in the content streaming system can be operated by a distributed server, and in this case, the data received by each server can be distributed processed.

5A is an embodiment to which the present invention can be applied. FIG. 5a is a QT (QuadTree, QT), FIG. 5b is a BT (Binary Tree, BT), and FIG. 5c is a TT (Ternary Tree, TT). ) FIG. 5d is a diagram for explaining a block division structure by AT (Asymmetric Tree, AT).

In video coding one block can be divided based on QT. Also, one subblock divided by QT can be recursively further divided using QT. Leaf blocks that are no longer QT divided can be divided by at least one of BT, TT or AT methods. The BT may have two forms of division, horizontal BT (2NxN, 2NxN) and vertical BT (Nx2N, Nx2N). The TT may have two forms of division: horizontal TT (2Nx1 / 2N, 2NxN, 2Nx1 / 2N) and vertical TT (1/2Nx2N, Nx2N, 1/2Nx2N). AT is horizontal-up AT (2Nx1 / 2N, 2Nx3 / 2N), horizontal-down AT (2Nx3 / 2N, 2Nx1 / 2N), vertical-left AT (1/2Nx2N, 3 / 2Nx2N), vertical-right AT (3). It can have four forms of division (/ 2Nx2N, 1/2Nx2N). Each BT, TT, AT can be recursively subdivided using BT, TT, AT.

FIG. 5a shows an example of QT division. Block A can be divided into four sub-blocks (A0, A1, A2, A3) by QT. The sub-block A1 can be divided into four sub-blocks (B0, B1, B2, B3) again by QT.

FIG. 5b shows an example of BT division. Block B3, which is no longer divided by QT, can be divided into vertical BT (C0, C1) or horizontal BT (D0, D1). Like block C0, each subblock can be recursively subdivided in the form of a horizontal BT (E0, E1) or vertical BT (F0, F1).

FIG. 5c shows an example of TT division. Block B3, which is no longer divided by QT, can be divided into vertical TT (C0, C1, C2) or horizontal TT (D0, D1, D2). Like block C1, each subblock can be recursively subdivided in the form of horizontal TT (E0, E1, E2) or vertical TT (F0, F1, F2).

FIG. 5d shows an example of AT division. Block B3, which is no longer divided by QT, can be divided into vertical AT (C0, C1) or horizontal AT (D0, D1). Like block C1, each subblock can be recursively subdivided in the form of horizontal AT (E0, E1) or vertical TT (F0, F1).

On the other hand, BT, TT, and AT division can be used together for division. For example, a subblock divided by BT can be divided by TT or AT. Further, the subblock divided by TT can be divided by BT or AT. The sub-block divided by AT can be divided by BT or TT. For example, after the horizontal BT division, each sub-block can be divided into vertical BT, or after the vertical BT division, each sub-block can be divided into horizontal BT. In this case, the order of division is different, but the final division is the same.

In addition, when the blocks are divided, the order in which the blocks are searched can be defined in various ways. In general, searching from left to right, top to bottom, and searching for blocks means the order in which you decide whether to split additional blocks in each split subblock, or blocks. When is no longer divided, it can mean the coding order of each subblock, or the search order when a subblock refers to information in another adjacent block.

The conversion can be performed for each processing unit (or conversion block) divided by the division structure as shown in FIGS. 5a to 5d, and in particular, the conversion is divided into the row direction and the column direction. A matrix can be applied. According to embodiments of the present invention, other conversion types are used, depending on the length of the processing unit (or conversion block) in the row or column direction.

The transformation is applied to the register dual block, which decorrelates the register dual block to the maximum extent, concentrates the coefficients at low frequencies, and creates a zero tail at the tip of the block. Because. In JEM software, the transformation part contains two main functions (core transform) and secondary transform. The core transform consists of a family of DCT (discrete cosine transform) and DST (discrete sine transform) transforms that apply to all rows and columns of the resilient block. After that, the secondary transformation can be additionally applied to the top left corner of the output of the core transformation. Similarly, the inverse transformation of the order of the quadratic inverse transformation and the core inverse transformation can be applied. First, the quadratic inverse transformation can be applied to the upper left corner of the coefficient block. After that, the core inverse transformation is applied to the rows and columns of the output of the quadratic inverse transformation. A core transformation or an inverse transformation can be referred to as a primary transformation or an inverse transformation.

6 and 7 show, as an embodiment to which the present invention is applied, FIG. 6 shows a conversion and quantization unit (120/130) and an inverse quantization and inverse conversion unit (140/130) in the encoding device 100 of FIG. A schematic block diagram of 150) is shown, and FIG. 7 shows a schematic block diagram of the inverse quantization and inverse conversion unit (220/230) in the decoding apparatus 200.

Looking carefully at FIG. 6, the transform and quantize unit (120/130) may include a primary transform unit 121, a secondary transform unit 122 and a quantize unit 130. can. The inverse quantization and inverse transformation unit (140/150) can include an inverse quantization unit 140, an inverse secondary transform unit 151, and an inverse primary transform unit 152. ..

Looking carefully at FIG. 7, the inverse quantization and inverse transform units (220/230) are the inverse quantize unit 220, the inverse secondary transform unit 231 and the inverse primary transform unit. ) 232 can be included.

In the present invention, when the conversion is performed, the conversion can be performed in a plurality of steps. For example, as shown in FIG. 6, two stages of a primary transform and a secondary transform can be applied, or more transformation steps are used based on the algorithm. Here, the primary transform can be referred to as a core transform.

The primary conversion unit 121 can apply the primary conversion to the resilient signal, where the primary conversion can be defaulted to the table from the encoder and / or the decoder.

The secondary conversion unit 122 can apply the secondary conversion to the linearly converted signal, and here the secondary conversion can be defaulted to the table from the encoder and / or the decoder.

In one embodiment, a non-separable secondary transform (NSST) can be conditionally applied as a secondary transform. For example, NSST may only be applied in the case of in-screen prediction blocks and may have a conversion set applicable for each prediction mode group.

Here, the prediction mode group can be set based on the symmetry of the prediction direction. For example, the prediction mode 52 and the prediction mode 16 are symmetrical based on the prediction mode 34 (diagonal direction), so that one group can be formed and the same transform set can be applied. At this time, when applying the conversion of the prediction mode 52, it is applied after transposing the input data, because the conversion set is the same as that of the prediction mode 16.

On the other hand, in the case of the planner mode (Planar mode) and the DC mode (DC mode), since there is no directional symmetry, each has a conversion set, and the conversion set can be composed of two conversions. .. For the remaining directional modes, each conversion set can consist of three transformations.

The quantization unit 130 can perform quantization on the second-order converted signal.

The inverse quantization and inverse transformation unit (140/150) reversely executes the process described above, and omits duplicate explanations.

FIG. 7 shows a schematic block diagram of the inverse quantization and inverse transformation unit (220/230) in the decoding apparatus 200.

Referring to FIG. 7, the inverse quantization and inverse transformation unit (220/230) includes an inverse quantization unit 220, an inverse secondary transform unit 231 and an inverse primary transform unit 232. Can include.

The inverse quantization unit 220 acquires a conversion coefficient from the entropy-decoded signal by using the quantization step size information.

In the inverse quadratic transformation unit 231, the inverse quadratic transformation is executed for the conversion coefficient. Here, the inverse quadratic transformation indicates the inverse transformation of the quadratic transformation described with reference to FIG.

The inverse primary conversion unit 232 performs inverse primary conversion on the signal (or block) that has been inverted secondary converted, and acquires a residual signal. Here, the inverse primary transformation shows the inverse transformation of the primary transform described with reference to FIG.

In addition to DCT-2 and 4x4 DST-4 applied to HEVC, adaptive multiple transform or explicit multiple transform (AMT or EMT) techniques are applied to the inter- and intra-encoded blocks. Used for resilient coding. Conversions in HEVC and a number of selected conversions from other DCT / DST families are used. The transformation matrices newly introduced by JEM are DST-7, DCT-8, DST-1 and DCT-5. Table 1 below shows the basis functions of the selected DST / DCT.

EMT can be applied to CUs that are less than 64 or have the same width and height, and whether EMT is applied can be controlled by the CU level flag. If the CU level flag is 0, then DCT-2 is applied to the CU to encode the residual. For the Luma coding block in the CU to which the EMT is applied, two additional flags are signaled to identify the horizontal and vertical transformations used. Like HEVC, JEM's block residency can be coded in conversion skip mode. For intra-residual coding, other resilience statistics of other intra-prediction modes use the process of selecting mode-dependent conversion candidates. The three conversion subsets are defined as shown in Table 2 below, and the conversion subset is selected based on the intra prediction mode as shown in Table 3.

Along with the subset concept, a subset of transformations are first identified based on Table 2 by using the intra-prediction mode of the CU with a CU-level EMT_CU_flag of 1. Hereafter, for each of the horizontal (EMT_TU_horizontal_flag) and vertical (EMT_TU_vertical_flag) transformations, one of the two conversion candidates within the subset of confirmed transformations is based on explicit signaling with flags, based on Table 3. Is selected.

Table 4 shows a transform configuration group to which AMT (adaptive multiple transform) is applied as an embodiment to which the present invention is applied.

Looking carefully at Table 4, the transform configuration group is determined based on the prediction mode, and the number of groups can be 6 (G0 to G5) in total. Then, G0 to G4 correspond to the case where the intra prediction is applied, and G5 indicates the combination of transformations (or the transformation set, the transformation combination set) applied to the residual block generated by the inter-prediction.

One combination of transformations is a horizontal transform (or row transform) applied to the row of the 2D block and a vertical transoform applied to the column (vertical transoform). Or it can be done by column transform.

Here, each group of conversion settings can have four conversion combination candidates. The four conversion combination candidates can be selected or determined via the index of 0 to 3 conversion combinations, and the index of the conversion combination from the encoder to the decoder can be encoded and transferred.

In one embodiment, the residual data (or residual signal) acquired via intra-prediction can have different statistical characteristics depending on the intra-prediction mode. Therefore, instead of the general cosine transform for each intra prediction as shown in Table 4, another transform can be applied. In the present specification, the conversion type can be expressed as, for example, DCT-Type 2, DCT-II, DCT-2.

A careful look at Table 4 reveals the use of 35 intra-prediction modes and the use of 67 intra-prediction modes. A combination of multiple transformations can be applied to each transformation setting group divided by the columns of each intra-prediction mode. For example, a combination of a plurality of conversions can be composed of a combination of four (row direction conversion, column direction conversion). As a specific example, in group 0, DST-7 and DCT-5 can be applied to all of the row (horizontal) direction and the column (vertical) direction, and a total of four combinations are possible.

Since a total of four transform kernel combinations can be applied to each intra prediction mode, the index of the transform combination for selecting one of them is transferred for each transform unit. be able to. In the present specification, the index of the conversion combination can be referred to as an AMT index and can be expressed by amt_idx.

In addition to the conversion kernels presented in Table 4, DCT-2 may be optimal in all of the row and column directions due to the characteristics of the residual signal. Therefore, by defining the AMT flag for each coding unit, the conversion can be applied adaptively. Here, if the AMT flag is 0, DCT-2 is applied in all the row and column directions, and if the AMT flag is 1, the AMT index is used to select or determine one of the four combinations. can do.

In one embodiment, if the AMT flag is 0 and the number of conversion coefficients in one conversion unit is less than 3, the conversion kernel in Table 4 is not applied and DST-7 is applied in all row and column directions. Can be

In one embodiment, the value of the conversion coefficient is parsed (analyzed) first, and if the number of conversion coefficients is smaller than 3, the AMT index is not analyzed and the additional information transfer amount is increased by applying DST-7. Can be reduced.

In one embodiment, AMT can only be applied if the total width and height of the conversion unit is 32 or less.

In one embodiment, Table 4 can be based on off-line training.

In one embodiment, the AMT index can be defined as one index that can simultaneously point to a combination of horizontal and vertical transformations. Alternatively, the AMT index can be defined by another horizontal and vertical conversion index.

FIG. 8 is a flowchart showing a process in which AMT (adaptive multiple transform) is executed.

In the present specification, an embodiment of a separable transform in which the transformation is applied separately in the horizontal direction and the vertical direction is basically described, but the combination of transformations is non-separable. It can also be configured as a non-separable transform.

Alternatively, a combination of transformations can be configured for a mixture of separable and non-separable transformations. In this case, if non-separable transformations are used, row / column transformation selections and horizontal / vertical transformation selections are no longer necessary, and only when separable transformations are selected, the transformation combinations in Table 4 are used. Can be

Further, the method proposed in the present specification can be applied regardless of the primary transformation or the secondary transformation. That is, there is no restriction that it must be applied to only one of the two, and it can be applied to both. Here, the primary transformation can mean the transformation to transform the residual block first, and the secondary transformation is the transformation to apply the transformation to the block generated as a result of the primary transformation. Can mean.

First, the encoding device 100 can determine the conversion group corresponding to the current block (S805). Here, the conversion group can mean the conversion group in Table 4, but the present invention is not limited to this, and can be composed of a combination of other conversions.

The encoding device 100 can perform conversions on the combination of candidate conversions used in the conversion group (S810). As a result of the conversion execution, the encoding device 100 can determine or select a combination of conversions having the minimum RD (rate distortion) cost (S815). The encoding device 100 can encode the index of the conversion combination corresponding to the selected conversion combination (S820).

FIG. 9 is a flowchart showing the decoding process in which AMT is executed.

First, the decoding device 200 can determine the conversion group for the current block (S905). The decoding device 200 can analyze the index of the conversion combination, and the index of the conversion combination can correspond to any one of the plurality of conversion combinations in the conversion group (S910). The decoding device 200 can derive a conversion combination corresponding to the index of the conversion combination (S915). Here, the conversion combination can mean the conversion combination described in Table 4, but the present invention is not limited thereto. That is, it is possible to configure by combining other conversions.

The decoding device 200 can perform inverse transformation on the current block based on the transformation combination (S920). If the transformation combination consists of a row transformation and a column transformation, the row transformation can be applied first, and then the column transformation can be applied. However, the present invention is not limited to this, and if it is applied in reverse or is composed of non-separable conversion, the direct non-separable conversion can be applied.

On the other hand, as another embodiment, the process of determining the transformation group and the process of analyzing the index of the transformation combination can be executed at the same time.

According to the embodiment of the present invention, the above-mentioned term "AMT" can be redefined as "MTS (multiple transform set or multiple transform selection)". The MTS-related syntax and semantics described below are defined in the VVC (versatile video coding) standard document JVET-K1001-v4.

In the embodiment of the present invention, two MTS candidates for the directional mode and four MTS candidates for the non-directional mode are used as described below.

A) Non-directional mode (DC, planner)

When the MTS index is 0, DST-7 is used for horizontal and vertical conversion.

When the MTS index is 1, DST-7 is used for vertical conversion and DCT-8 is used for horizontal conversion.

When the MTS index is 2, DCT-8 is used for vertical conversion and DST-7 is used for horizontal conversion.

When the MTS index is 3, DCT-8 is used for horizontal and vertical conversion.

B) Modes that belong to the horizontal group mode

When the MTS index is 0, DST-7 is used for horizontal and vertical conversion.

When the MTS index is 1, DCT-8 is used for vertical conversion and DST-7 is used for horizontal conversion.

C) A mode that belongs to the vertical group mode

When the MTS index is 0, DST-7 is used for horizontal and vertical conversion.

When the MTS index is 1, DST-7 is used for vertical conversion and DCT-8 is used for horizontal conversion.

Here (in VTM 2.0, where 67 modes are used), the horizontal group mode includes the 2nd to 34th intra prediction modes and the vertical mode includes the 35th to 66th intra prediction modes.

In another embodiment of the invention, three MTS candidates are used for all intra prediction modes.

When the MTS index is 0, DST-7 is used for horizontal and vertical conversion.

When the MTS index is 1, DST-7 is used for vertical conversion and DCT-8 is used for horizontal conversion.

When the MTS index is 2, DCT-8 is used for vertical conversion and DST-7 is used for horizontal conversion.

In another embodiment of the invention, two MTS candidates are used for the directional prediction mode and three MTS candidates are used for the non-directional prediction mode.

A) Non-directional mode (DC, planner)

When the MTS index is 0, DST-7 is used for horizontal and vertical conversion.

When the MTS index is 1, DST-7 is used for vertical conversion and DCT-8 is used for horizontal conversion.

When the MTS index is 2, DCT-8 is used for vertical conversion and DST-7 is used for horizontal conversion.

B) Prediction mode corresponding to the horizontal group mode

When the MTS index is 0, DST-7 is used for horizontal and vertical conversion.

When the MTS index is 1, DCT-8 is used for vertical conversion and DST-7 is used for horizontal conversion.

C) Prediction mode corresponding to the vertical group mode

When the MTS index is 0, DST-7 is used for horizontal and vertical conversion.

When the MTS index is 1, DST-7 is used for vertical conversion and DCT-8 is used for horizontal conversion.

In another embodiment of the invention, one MTS candidate (eg, DST-7) is used for all intramodes. In this case, the encoding time can be reduced to 40% with minor coding loss. In addition, one flag is used to indicate between DCT-2 and DST-7.

FIG. 10 is a flowchart showing an inverse transformation process based on MTS based on the embodiment of the present invention.

The decoding device 200 to which the present invention is applied can acquire sps_mts_intra_enabled_flag or sps_mts_inter_enabled_flag (S1005). Here, sps_mts_intra_enabled_flag indicates whether cu_mts_flag is present in the resilient coding syntax of the intracoding unit. For example, if sps_mts_intra_enabled_flag = 0, cu_mts_flag does not exist in the intracoding unit's resilient coding syntax, and if sps_mts_intra_enabled_flag = 1, cu_mts_flag exists in the intracoding unit's resilient coding syntax. And sps_mts_inter_enabled_flag indicates whether cu_mts_flag is present in the redundant coding syntax of the intercoding unit. For example, if sps_mts_inter_enabled_flag = 0, cu_mts_flag does not exist in the intercoding unit's resilient coding syntax, and if sps_mts_inter_enabled_flag = 1, cu_mts_flag exists in the intercoding unit's resilient coding syntax.

The decoding device 200 can acquire the cu_mts_flag based on sps_mts_intra_enabled_flag or sps_mts_inter_enabled_flag (S1010). For example, when sps_mts_intra_enabled_flag = 1 or sps_mts_inter_enabled_flag = 1, the decoding device 200 can acquire cu_mts_flag. Here, cu_mts_flag indicates whether MTS is applied to the resilient sample of the Luma conversion block. For example, if cu_mts_flag = 0, MTS is not applied to the register dual sample of the Luma conversion block, and if cu_mts_flag = 1, MTS is applied to the register dual sample of the Luma conversion block.

The decoding device 200 can acquire mts_idx based on cu_mts_flag (S1015). For example, when cu_mts_flag = 1, the decoding device 200 can acquire mts_idx. Here, mts_idx indicates which transformation kernel is currently applied to the Luma-residual sample along the horizontal and / or vertical orientation of the transformation block.

For example, for mts_idx, at least one of the embodiments described herein can be applied.

The decoding device 200 can guide the conversion kernel corresponding to mts_idx (S1020). For example, the conversion kernel corresponding to mts_idx can be defined separately for horizontal conversion and vertical conversion.

As an example, when MTS is applied to the current block (ie, cu_mts_flag = 1), the decoding device 200 can configure MTS candidates based on the intra prediction mode of the current block. In this case, the decoding flowchart of FIG. 10 may further include a step of configuring the MTS candidate. Then, the decoding device 200 can determine the MTS candidate currently applied to the block by using mts_idx from the configured MTS candidates.

As another example, horizontal and vertical transformations can be applied with different transformation kernels. However, the present invention is not limited to this, and the same conversion kernel may be applied to the horizontal conversion and the vertical conversion.

Then, the decoding device 200 can perform inverse transformation based on the conversion kernel (S1025).

Further, in this document, MTS can also be expressed as AMT or EMT, and similarly, it can be expressed as mts_idx figure AMT_idx, EMT_idx, AMT_TU_idx EMT_TU_idx, etc., and the present invention is not limited to such expressions. ..

Further, in the present invention, the case where the MTS is applied based on the MTS flag and the case where the MTS is not applied will be described separately, but the present invention is not limited to such an expression. For example, whether or not to apply MTS is whether to use another conversion type (or conversion kernel) other than the predetermined specific conversion type (which may be referred to as a basic conversion type, a default conversion type, etc.). It can have the same meaning as whether or not. If MTS is applied, another conversion type other than the basic conversion type (eg, any one or a combination of two or more conversion types of multiple conversion types) is used for the conversion and the MTS is applied. If not, the basic conversion type can be used for the conversion. In one embodiment, the basic conversion type can be set (or defined) in DCT2.

As an example, the MTS flag syntax, which indicates whether or not the MTS is applied to the current conversion block, and the MTS index syntax, which indicates the conversion type applied to the current block when the MTS is applied, are individually output from the encoder. A syntax that may be sent to the decoder and, as another example, includes whether or not the MTS is currently applied to the conversion block and, if the MTS is applied, the conversion type that is applied to the current block. For example, the MTS index) may be transmitted from the encoder to the decoder. That is, in the latter embodiment, a syntax (or thin) indicating the conversion type currently applied to the conversion block (or unit) within the overall conversion type group (or conversion type set) including the basic conversion type described above. Tax elements) may be sent from the encoder to the decoder.

Therefore, the syntax (MTS index) indicating the conversion type currently applied to the conversion block despite the representation can include information on whether or not to apply the MTS. In other words, in the latter embodiment, only the MTS index is signaled without the MTS flag. Therefore, in this case, it can be interpreted that DCT2 is included in MTS, but in the present invention, MTS is applied when DCT2 is applied. It may be described as not, and nevertheless, the technical scope regarding MTS is not limited to the content of the definition.

FIG. 11 is a block diagram of an apparatus that performs decoding based on MTS based on an embodiment of the present invention.

The decoding device 200 to which the present invention is applied can include a sequence parameter acquisition unit 1105, an MTS flag acquisition unit 1110, an MTS index acquisition unit 1115, and a conversion kernel induction unit 1120.

The sequence parameter acquisition unit 1105 can acquire sps_mts_intra_enabled_flag or sps_mts_inter_enabled_flag. Here, sps_mts_intra_enabled_flag indicates whether cu_mts_flag exists in the resilient coding syntax of the intracoding unit, and sps_mts_inter_enabled_flag indicates whether cu_mts_flag exists in the resilient coding syntax of the intercoding unit. As a specific example, the description related to FIG. 10 can be applied.

The MTS flag acquisition unit 1110 can acquire cu_mts_flag based on sps_mts_intra_enabled_flag or sps_mts_inter_enabled_flag. For example, when sps_mts_intra_enabled_flag = 1 or sps_mts_inter_enabled_flag = 1, the MTS flag acquisition unit 1110 can acquire cu_mts_flag. Here, cu_mts_flag indicates whether MTS is applied to the resilient sample of the Luma conversion block. As a specific example, the description related to FIG. 10 can be applied.

The MTS index acquisition unit 1115 can acquire mts_idx based on cu_mts_flag. For example, when cu_mts_flag = 1, the MTS index acquisition unit 1115 can acquire mts_idx. Here, mts_idx indicates which transformation kernel is applied to the Luma-residual sample along the horizontal and / or vertical direction of the current transformation block. As a specific example, the description of FIG. 10 can be applied.

The conversion kernel induction unit 1120 can induce the conversion kernel corresponding to mts_idx. Then, the decoding device 200 can perform the inverse transformation based on the derived transformation kernel.

A mode-dependent non-separable secondary transform (MDNSST) is introduced. To maintain low complexity, MDNSST is applied only to the post-low frequency coefficients after the first-order conversion. Further, the non-separable transform mainly applied to the low frequency coefficient can be called LFNST (low frequency non-separable transform). If the width (width, W) and height (height, H) of the conversion coefficient block are all 8 or more, the 8x8 non-separable secondary conversion is applied to the region of 8x8 on the upper left side of the conversion coefficient block. Otherwise, if the width or height is less than 8, a 4x4 non-separable quadratic transformation is applied and the 4x4 non-separable quadratic transformation is the upper left min (8, W) x min (8, W) of the conversion factor block. It can be executed in H). Here, min (A, B) is a function that outputs a smaller value among A and B. Further, WxH indicates the size of the block, W indicates the width, and H indicates the height.

In one embodiment, there can be a total of 35x3 non-separable secondary transformations for 4x4 and 8x8 block sizes, where 35 is the number of transformation sets specified by the intra prediction mode and 3 is for each prediction mode. The number of NSST candidates. The mapping from the intra prediction mode to the transformation set can be defined as shown in Table 5 below.

Further, according to the embodiment of the present invention, in the four non-separable conversion sets, depending on the intra prediction mode. The NSST index (NSST idx) can be coded to indicate the conversion kernel. If NSST is not applied, an NSST index with a value of 0 can be signaled.

12 and 13 are encoding / decoding flowcharts to which a secondary transformation is applied as an embodiment to which the present invention is applied.

In JEM, secondary transformation (MDNSST) does not apply to blocks coded in conversion skip mode. If the MDNSST index is signaled to the CU and is non-zero, the MDNSST is not used for blocks of components coded in conversion skip mode within the CU. The overall coding structure, including coefficient encoding and NSST index coding, is shown in FIGS. 12 and 13. The CBF (coded block flag) is encoded to determine whether to code the coefficients and NSST. In FIGS. 12 and 13, the CBF flag can indicate a Lumablock cbf flag (cbf_luma flag) or a chromablock cbf flag (cbf_cb flag or cbf_cr flag). The conversion factor when the CBF flag is 1 is coded.

Referring to FIG. 12, the encoding device 100 confirms whether the CBF is 1 (S1205). If the CBF is 0, the encoding device 100 does not perform conversion factor encoding and NSST index encoding. When the CBF is 1, the encoding device 100 encodes the conversion coefficient (S1210). After that, the encoding device 100 determines whether or not to perform NSST index coding (S1215), and then performs NSST index coding (S1220). If NSST index coding is not applied, the encoding device 100 can complete the conversion procedure to the state where NSST is not applied and perform subsequent steps (eg, quantization).

Referring to FIG. 13, the decoding device 200 confirms whether or not the CBF is 1 (S1305). When the CBF is 0, the decoding device 200 has not executed the decoding of the conversion coefficient and the NSST index decoding. When the CBF is 1, the decoding device 200 decodes the conversion coefficient (S1310). After that, the decoding device 200 determines whether or not to perform NSST index coding (S1315), and analyzes the NSST index (S1320).

NSST does not apply to the entire block (TU in the case of HEVC) to which the primary transformation is applied, but can be applied to the upper left 8x8 region or 4x4 region. As an example, if the block size is 8x8 or more, 8x8 NSST can be applied, and if it is less than 8x8, 4x4 NSST can be applied. Further, when 8x8 NSST is applied, 4x4 NSST can be applied every 4x4 block. The 8x8 NSST and the 4x4 NSST can all be determined according to the configuration of the conversion set described above, and only the non-separable conversion 8x8 NSST has 64 input data and 64 output data, and the 4x4 NSST is It can have 16 inputs and 16 outputs.

14 and 15 are embodiments to which the present invention is applied, in which FIG. 14 shows a diagram for explaining Givens rotation, and FIG. 15 is composed of a Givens rotation layer and a permutation. The configuration of one round in the 4x4 NSST is shown.

All 8x8 NSST and 4x4 NSST can be configured with a hierarchical combination of Givens rotations. The matrix corresponding to one Given rotation is the same as the mathematical formula 1, and the matrix product is the same as FIG. 14 when expressed graphically.

In FIG. 14, tm and tn output by Givens rotation can be calculated as in mathematical formula 2.

Since one Givens rotation rotates two data as shown in FIG. 14, 32 pieces each for processing 64 pieces of data (in the case of 8x8 NSST) or 16 pieces of data (in the case of 4x4 NSST), respectively. Or 8 Givens rotations are required. Therefore, a bundle of 32 or 8 Given rotations can form a Given rotation layer. As shown in FIG. 15, the output data of one Given Rotation Layer is transmitted to the input data of the next Given Rotation Layer via substitution (shuffle). The patterns to be substituted as shown in FIG. 15 are regularly defined, and in the case of 4x4 NSST, the four Givens rotation layers and the corresponding substitutions form one round. The 4x4 NSST is performed in two rounds and the 8x8 NSST is performed in four rounds. Different rounds use the same substitution pattern, but the applied given rotation angles are different. Therefore, it is necessary to store the angular data of all Givens rotations that make up each transmutation.

Finally, one more substitution is performed on the data output through the Givens rotation layer in the final stage, and the information on the substitution is stored separately for each conversion. At the end of the forward NSST, the substitution is made, and for the inverse NSST, the reverse substitution is applied first.

The reverse NSST replaces the Givens rotation layer applied in the forward NSST in the reverse order, and rotates by taking a negative (-) value for each Given rotation angle.

RST (Reduced secondary transform)

FIG. 16 shows the operation of RST as an embodiment to which the present invention is applied.

Assuming that the orthogonal matrix indicating one transformation has the form of NxN, RT (reduced transform) leaves only R out of N transformed basis vectors (R <N). The matrix of forward RT that produces the transformation coefficients can be defined as in mathematical equation 3.

Since the matrix of the reverse RT is a transpose matrix of the forward RT matrix, it may be the same as in FIGS. 16a and 16b if the application of the forward RT and the reverse RT is illustrated.

The RT applied to the 8x8 block at the upper left corner of the block with the conversion factor to which the linear transformation is applied can be called the 8x8 RST. In mathematical formula 3, when the value of R is set to 16, the forward 8x8 RST has the form of a 16x64 matrix and the reverse 8x8 RST has the form of 64x16. Further, the configuration of the conversion set as shown in Table 5 can be applied to 8x8 RST. That is, 8x8 RST can be determined based on the conversion set according to the intra prediction mode as shown in Table 5. Since one conversion set consists of two or three transformations depending on the intra prediction mode, one of the maximum four transformations is selected, including the case where the secondary transformation is not applied. (One transformation can be an identity matrix). When 0, 1, 2, and 3 indexes are assigned to each of the four conversions, the syntax element corresponding to the NSST index is signaled for each block of conversion coefficients. Conversion can be specified. For example, index 0 can be assigned as an identity matrix, i.e., when no quadratic transformation is applied. In conclusion, for the 8x8 upper left block via the NSST index, 8x8 NSST may be specified according to JEM NSST, and 8x8 RST can be specified according to the RST configuration.

FIG. 17 is a diagram showing a process of performing reverse scans from the 64th to the 17th based on the reverse scan order as an embodiment to which the present invention is applied.

When a forward 8x8 RST such as math formula 3 is applied, 16 valid conversion coefficients are generated, so that the 64 input data constituting the 8x8 area is reduced to 16 output data. From the viewpoint of the two-dimensional region, the effective conversion coefficient is satisfied only in the region of about 1/4. Therefore, by applying the forward 8x8 RST, the acquired 16 output data are filled in the upper left region of FIG.

In FIG. 17, the upper left 4x4 region is the ROI (region of interest) region where the effective conversion coefficient is satisfied, and the remaining region is empty. The vacant area can be filled with a value of 0 by default. If a valid non-zero conversion factor is found in addition to the ROI region of FIG. 17, it is certain that the 8x8 RST will not be applied and the relevant coding may be omitted in the NSST index. Conversely, if a non-zero conversion coefficient is not found outside the ROI region of FIG. 17 (when 8x8 RST is applied, when the region other than ROI is filled with 0), 8x8 RST may have been applied. Therefore, the NSST index can be coded. Such conditional NSST index coding is non-zero and can be performed after the resilient coding process as it requires checking for the presence or absence of conversion factors.

FIG. 18 shows an example of an encoding flowchart using a single transform indicator as an embodiment to which the present invention is applied.

In an embodiment of the invention, a single transform indicator (STI) is introduced. Instead of using two transforms in sequence (first-order transform and second-order transform), a single transform can be applied when the single transform indicator is activated (STI coding == 1). .. Here, a single conversion can be any kind of conversion. For example, a single transformation can be a separable transformation or a non-separable transformation. The single transformation can be a transformation approximated from the non-separable transformation. The single conversion index (ST_idx in FIG. 18) can be signaled when the single conversion indicator is activated. Here, the single conversion index can indicate the conversion corresponding to the conversion to be applied from among the available conversion candidates.

Referring to FIG. 18, the encoding device 100 determines whether the CBF is 1 (S1805). If the CBF is 1, the encoding device 100 determines whether STI coding is applied (S1810). When STI coding is applied, the encoding device 100 encodes the STI index (STI_Idx) (S1845) and codes the conversion coefficient (S1850). If no STI coding is applied, the encoding device 100 encodes a flag (EMT_CU_Flag) indicating whether EMT (or MTS) is applied at the CU level (S1815). After that, the encoding device 100 codes the conversion coefficient (S1820). After that, it is determined whether or not EMT is applied to the encoding device 100 conversion unit (TU) (S1825). When EMT is applied to the TU, the encoding device 100 encodes the index (EMT_TU Idx) of the linear transformation applied to the TU (S1830). After that, the encoding device 100 determines whether or not NSST is applied (S1835). When NSST is applied, the encoding device 100 encodes an index (NSST_Idx) indicating the applied NSST (S1840).

In one example, when the single conversion coding condition is satisfied / activated (eg, STI_coding == 1), the single conversion index (ST_Idx) is not signaled and can be implicitly derived. .. ST_idx can be implicitly determined based on the block size and intra-prediction mode. Here, ST_idx can indicate the transformation (or transformation kernel) applied to the current transformation block.

The single conversion indicator can be activated when one or more of the following conditions are satisfied (STI_coding == 1).

The block size corresponds to a predetermined value, such as 4 or 8.

Block width == Block height (square block)

It is an intra prediction mode of any one of predetermined modes such as DC or planner.

In another example, the STI coding flag can be signaled to indicate whether a single transformation is applied. The STI coding flag can be signaled based on the STI coding value and the CBF. For example, the STI coding flag has a CBF of 1 and can be signaled when STI coding is activated. In addition, the STI coding flag can be signaled conditionally, taking into account the block size, block shape (square or non-square block), or intra-prediction mode.

Of the coefficient coding, ST_idx can be determined after the coefficient coding because the acquired information is used. In one example, ST_idx can be implicitly determined based on the block size, the intra prediction mode, and the number of non-zero coefficients. In another example, ST_idx can conditionally encode / decode based on block size and / or block shape and / or intra-prediction mode, and / or number of non-zero coefficients. In another example, ST_idx signaling can be omitted depending on the distribution of non-zero coefficients (ie, the position of the non-zero coefficients). In particular, if a non-zero coefficient is found in a region other than the upper left 4x4 region, ST_idx signaling can be omitted.

FIG. 19 shows an example of an encoding flowchart using a unified transform indicator (UTI) as an embodiment to which the present invention is applied.

In the embodiment of the present invention, a unified conversion indicator is introduced. The UTI includes a primary conversion indicator and a secondary conversion indicator.

Referring to FIG. 19, the encoding device 100 determines whether the CBF is 1 (S1905). If the CBF is 1, the encoding device 100 determines whether UTI coding is applied (S1910). When UTI coding is applied, the encoding device 100 encodes the UTI index (UTI_Idx) (S1945) and codes the conversion factor (S1950). If no UTI coding is applied, the encoding device 100 encodes a flag (EMT_CU_Flag) indicating whether EMT (or MTS) is applied at the CU level (S1915). After that, the encoding device 100 encodes the conversion coefficient (S1920). After that, it is determined whether or not EMT is applied to the encoding device 100 conversion unit (TU) (S1925). When EMT is applied to the TU, the encoding device 100 encodes the index (EMT_TU Idx) of the linear transformation applied to the TU (S1930). After that, the encoding device 100 determines whether or not NSST is applied (S1935). When NSST is applied, the encoding device 100 encodes an index (NSST_Idx) indicating the applied NSST (S1940).

The UTI can be encoded for each predetermined unit (CTU or CU).

The UTI coding mode can depend on the following conditions.

Block size

Block morphology

Intra prediction mode

How to derive / extract the core transformation index from the UTI is predefined. How to derive / extract the secondary transformation index from the UTI is predefined.

The UTI syntax structure is used selectively. The UTI can depend on the CU (or TU) size. For example, a smaller CU (TU) can have a relatively narrow range of UTI indexes. In one example, if a predefined condition (eg, the block size is less than a predefined threshold) is satisfied, the UTI can only indicate the core transformation index.

In another example, the UTI index can be treated as a core transformation index if the secondary transformation is not instructed to be used (eg, if the secondary transformation index == 0 or the secondary transformation has already been determined). Similarly, if the core conversion index is known, the UTI index can be treated as a secondary conversion index. In particular, a predetermined core transformation can be used, taking into account the intra-prediction mode and block size.

20a and 20b show another example of an encoding flowchart using UTI as an embodiment to which the present invention is applied.

In another example, the conversion coding structure uses UTI index coding as shown in FIGS. 20a and 20b. Here, the UTI index can be coded before or after the coefficient coding.

Referring to FIG. 20a, the encoding device 100 confirms whether the CBF is 1 (S2005). If the CBF is 1, the encoding device 100 codes the UTI index (UTI_Idx) (S2010) and codes the conversion coefficient (S2015).

Referring to FIG. 20b, the encoding device 100 confirms whether the CBF is 1 (S2055). If the CBF is 1, the encoding device 100 performs coding of the conversion factor (S2060) and codes the UTI index (UTI_Idx) (S2065).

In another embodiment of the invention, data hiding and implicit coding methods of conversion directives are introduced. Here, the conversion specifier includes ST_idx, UTI_idx, EMT_CU_Flag, EMT_TU_Flag, NSST_idx and indexes related to the conversion used to indicate the conversion kernel. The conversion directive described above is not signaled and the information can be inserted into the coefficient encoding process (can be extracted within the coefficient coding process). The coefficient encoding process can include the following parts:

− Last x position (Last_position_x), last y position (Last_position_y)

− Group flag

− Significance map

-Flag indicating whether it is larger than 1 (Greater_than_1_flag)

-Flag indicating whether it is larger than 2 (Greater_than_2_flag)

− Remaining level coding

-Sign coding

For example, conversion directive information can be inserted into one or more of the coefficient coding processes described above. Here are some things that can be considered together to insert conversion directive information:

-Pattern of Sign coding

− The absolute value of remaining level

The number of Greater_than_1_flag that indicates whether it is greater than -1

-The value of Last_position_X and Last_position_Y

The above-mentioned data hiding method can be considered conditionally. For example, the data hiding method may depend on the number of non-zero coefficients.

In yet another example, NSST_idx and EMT_idx can be dependent. For example, when EMT_CU_flag is 0 (or 1), NSST_idx may not be 0. In this case, NSST_idx-1 can signal instead of NSST_idx.

In another embodiment of the invention, the mapping of NSST conversion sets based on the intra prediction mode is introduced as shown in Table 7 below. As mentioned above, the following description will focus on NSST as an example of non-separable conversion, but other known terms for non-separable conversion (eg, LFNST) can be used. For example, the NSST set and the NSST index are replaced with the LFNST set and the LFNST index. Also, the RST described in this document is a square that applies to at least a portion of the transformation block (the upper left 4x4, 8x8 region or the remaining 8x8 block excluding the right-lower 4x4 region). As an example of a non-separable transformation (eg, LFNST) using a non-square transformation matrix with a reduced input length and / or a reduced output length in a non-separable transformation matrix, it is used in place of RST or LFNST. Be done.

NSST set numbers can be rearranged between 0 and 3 as shown in Table 8.

In the NSST conversion set, four conversion sets are used (instead of 35) to reduce the required memory space.

In addition, different numbers of conversion kernels are used for each (each) conversion set as follows:

Case A: Two available conversion kernels are used for each (each) conversion set and the NSST index range is 0-2. For example, if the NSST index is 0, the secondary transformation (secondary inverse transformation based on the decoder) may not be applied. If the NSST index is 1 or 2, a secondary transformation can be applied. The conversion set can include two conversion kernels, and one or two indexes can be mapped to the two conversion kernels.

Referring to Table 9, two conversion kernels are used for each 0 to 3 non-separable conversion (NSST or LFNST) set.

Case B: Two available conversion kernels are used for the 0th conversion set, and one conversion kernel is used for each of the remaining conversion sets. The NSST index that can be used for the 0th conversion set (DC, planner) is 0 to 2. However, the NSST index of the other modes (conversion sets 1, 2, and 3) is 0 to 1.

Referring to Table 10, two non-separable conversion kernels are set for the non-separable conversion (NSST) set corresponding to the 0th index, and the non-separable conversion (NSST) set corresponding to the 1st, 2nd and 3rd indexes. One non-separable conversion kernel is set for each.

Case C: One conversion kernel is used for each (each) conversion set, and the NSST index range is 0 to 1.

FIG. 21 shows an example of an encoding flowchart that executes conversion as an embodiment to which the present invention is applied.

The encoding device 100 performs a primary conversion on the resilient block (S2105). The primary transformation can be referred to as the core transformation. As an embodiment, the encoding device 100 can perform the primary conversion using the MTS described above. Further, the encoding device 100 can transfer the MTS index indicating the specific MTS from the MTS candidates to the decoding device 200. At this time, the MTS candidate can be configured based on the intra prediction mode of the current block.

The encoding device 100 determines whether or not to apply the secondary transformation (S2110). As an example, the encoding device 100 can determine whether or not to apply the secondary transformation based on the linearly transformed resilient conversion coefficient. For example, the secondary transformation can be NSST or RST.

The encoding device 100 determines the secondary conversion (S2115). At this time, the encoding device 100 can determine the secondary conversion based on the NSST (or RST) conversion set specified according to the intra prediction mode.

Further, as an example, the encoding device 100 can determine the region to which the secondary transformation is applied based on the size of the current block prior to the S2115 step.

The encoding device 100 performs the secondary transformation using the secondary transformation determined in the S2115 step (S2120).

FIG. 22 shows an example of a decoding flowchart that executes conversion as an embodiment to which the present invention is applied.

The decoding device 200 determines whether or not to apply the second-order inverse transformation (S2205). For example, the second-order inverse transformation can be NSST or RST. As an example, the decoding apparatus 200 can determine whether or not to apply the secondary inverse transformation based on the secondary transformation flag received from the encoding apparatus 100.

The decoding device 200 determines the second-order inverse transformation (S2210). At this time, the decoding device 200 can determine the second-order inverse transformation applied to the current block based on the NSST (or RST) transformation set specified according to the intra-prediction mode described above.

Further, as an example, the decoding device 200 can determine the region to which the second-order inverse transformation is applied based on the size of the current block prior to the S2210 step.

The decoding device 200 performs the second-order inverse transformation on the inverse-quantized residency block using the second-order inverse transformation determined in the S2210 step (S2215).

The decoding device 200 performs the first-order inverse transformation on the second-order inverse-converted register dual block (S2220). The first-order inverse transformation can be called the core inverse transformation. As an embodiment, the decoding device 200 can perform primary inverse transformation using the above-mentioned MTS. Also, as an example, the decoding device 200 can determine whether or not the MTS is currently applied to the block prior to the S2220 stage. In this case, the decoding flowchart of FIG. 22 may further include a step of determining whether MTS is applied.

As an example, if MTS is applied to the current block (ie, cu_mts_flag = 1), the decoder 200 can configure MTS candidates based on the current block's intra-prediction mode. In this case, the decoding flowchart of FIG. 22 may further include a step of configuring the MTS candidate. Then, the decoding device 200 can determine the primary inverse transformation applied to the current block by using mts_idx indicating a specific MTS among the configured MTS candidates.

FIG. 23 shows an example of a detailed block diagram of the conversion unit 120 in the encoding device 100 as an embodiment to which the present invention is applied.

The encoding device 100 to which the embodiment of the present invention is applied can include a primary conversion unit 2310, a secondary conversion applicability determination unit 2320, a secondary conversion determination unit 2330, and a secondary conversion unit 2340.

The primary conversion unit 2310 can perform the primary conversion on the resilient block. The primary transformation can be referred to as the core transformation. As an embodiment, the primary conversion unit 2310 can perform the primary conversion using the MTS described above. Further, the primary conversion unit 2310 can transfer the MTS index indicating the specific MTS from the MTS candidates to the decoding device 200. At this time, the MTS candidate can be configured based on the intra-prediction mode of the current block.

The secondary conversion applicability determination unit 2320 can determine whether or not to apply the secondary conversion. As an example, the secondary conversion applicability determination unit 2320 can determine whether or not to apply the secondary conversion based on the conversion coefficient of the linearly converted resilient block. For example, the secondary transformation can be NSST or RST.

The secondary conversion determination unit 2330 determines the secondary conversion. At this time, as described above, the secondary conversion determination unit 2330 can determine the secondary conversion based on the NSST (or RST) conversion set specified according to the intra prediction mode.

Further, as an example, the secondary transformation determination unit 2330 can also determine the region to which the secondary transformation is applied based on the size of the current block.

The secondary conversion unit 2340 can execute the secondary conversion using the determined secondary conversion.

FIG. 24 shows an example of a detailed block diagram of the inverse conversion unit 230 in the decoding device 200 as an embodiment to which the present invention is applied.

The decoding device 200 to which the present invention is applied includes a secondary inverse conversion applicability determination unit 2410, a secondary inverse conversion determination unit 2420, a secondary inverse conversion unit 2430, and a primary inverse conversion unit 2440.

The second-order inverse transformation applicability determination unit 2410 can determine whether or not to apply the second-order inverse transformation. For example, the second-order inverse transformation can be NSST or RST. As an example, the secondary inverse transformation applicability determination unit 2410 can determine whether or not to apply the secondary inverse transformation based on the secondary transformation flag received from the encoding device 100. As another example, the second-order inverse transformation applicability determination unit 2410 can also determine whether or not to apply the second-order inverse transformation based on the conversion coefficient of the resilient block.

The secondary inverse transformation determination unit 2420 can determine the secondary inverse transformation. At this time, the secondary inverse transformation determination unit 2420 can determine the secondary inverse transformation applied to the current block based on the NSST (or RST) transformation set specified according to the intra prediction mode.

Further, as an example, the secondary inverse transformation determination unit 2420 can determine the region to which the secondary inverse transformation is applied based on the size of the current block.

Further, as an example, the second-order inverse transformation unit 2430 can perform the second-order inverse transformation on the inverse-quantized residency block by using the determined second-order inverse transformation.

The primary inverse transformation unit 2440 can perform the primary inverse transformation on the register dual block that has been transformed into the secondary inverse. As an embodiment, the primary inverse conversion unit 2440 can perform the primary conversion using the MTS described above. Further, as an example, the primary inverse conversion unit 2440 can determine whether or not the MTS is currently applied to the block.

As an example, when MTS is applied to the current block (that is, cu_mts_flag = 1), the primary inverse transformant 2440 can configure MTS candidates based on the intra-prediction mode of the current block. Then, the primary inverse transformation unit 2440 can determine the primary transformation applied to the current block by using mts_idx indicating a specific MTS from the configured MTS candidates.

FIG. 25 shows a flowchart for processing a video signal as an embodiment to which the present invention is applied. The flowchart of FIG. 25 can be executed by the decoding device 200 or the inverse conversion unit 230.

First, the decoding device 200 can determine whether to apply the inverse non-separable conversion of the current block based on the non-separable conversion index and the width and height of the current block. For example, the decoding device 200 can determine to apply the non-separable conversion when the non-separable conversion index is not 0 and the width and height of the current block are each 4 or more. If the non-separable conversion index is 0 or the width or height of the current block is less than 4, the decoding device 200 can omit the non-separable conversion in the reverse direction and perform the primary conversion in the reverse direction. ..

In the step S2505, the decoding device 200 indicates a non-separable conversion set to be used for the non-separable conversion of the current block from among the predefined non-separable conversion sets based on the intra-prediction mode of the current block. Determine the set index. The non-separable conversion set index can be set to be assigned to each of the four conversion sets set according to the range of the intra prediction mode, as shown in Table 7 or Table 8. That is, as shown in Table 7 or Table 8, when the intra prediction mode is 0 to 1, the non-separable conversion set index is determined to be the first index value, and the intra prediction mode is 2 to 12 or 56 to 66. For example, if the non-separable conversion set index is determined to be the second index value and the intra prediction mode is 13 to 23 or 45 to 55, the non-separable conversion set index is determined to be the third index value and the intra prediction mode. If is 24 to 44, the non-separable conversion set index can be determined to be the fourth index value.

Here, the predefined non-separable conversion set can each include two conversion kernels, as shown in Table 9. The predefined non-separable conversion set can also include one or two conversion kernels, as in Table 10 or Table 11.

In the S2510 step, the decoding apparatus 200 uses a non-separable transformation matrix to select the transformation kernel specified by the non-separable transformation index of the current block among the transformation kernels included in the non-separable transformation set indicated by the non-separable transformation set index. decide. For example, two non-separable transformation kernels can be set for each of the index values of the non-separable transformation set index, and the decoding device 200 can be set as a non-separable transformation matrix kernel among the two transformation matrix kernels corresponding to the non-separable transformation set index. The non-separable transformation matrix can be determined based on the transformation kernel indicated by the separate transformation index.

In step S2515, the decoding apparatus 200 applies a non-separable transformation matrix to the upper left region of the current block, which is determined according to the width and height of the current block. For example, if the width and height of the current block are all 8 or more, the non-separable transformation is applied to the 8x8 area on the upper left side of the current block, and the width or height of the current block is less than 8. In this case, the non-separable transformation can be applied to the upper left 4x4 region of the current block. The size of the non-separable conversion can also be set to a size corresponding to 8x8 or 4x4 (e.g. 48x16, 16x16) corresponding to the region to which the non-separable conversion is applied.

Further, the decoding device 200 can apply the horizontal conversion and the vertical conversion to the current block to which the non-separable conversion is applied. Here, the horizontal and vertical transformations can be determined based on the prediction mode currently applied to the block and the MTS index for the selection of the transformation matrix.

In the following, a method of applying a combination of a primary transformation and a secondary transformation will be described. That is, in the embodiment of the present invention, a method for efficiently designing the transformation used for the primary transformation and the secondary transformation is proposed. Here, the method proposed in FIGS. 1 to 25 may be applied, and related and duplicated description will be omitted.

As mentioned above, the primary transformation indicates the transformation applied first to the residual block relative to the encoder. When a secondary transformation is applied, the encoder performs a secondary transformation on the primary transformed residual block. On the other hand, when the secondary transformation is applied, the secondary inverse transformation is performed prior to the primary inverse transformation with reference to the decoder. The decoder can derive a residual block by performing a first-order inverse transformation on a conversion coefficient block that has undergone a second-order inverse transformation.

Further, as described above, a non-separable transformation may be used as a secondary transformation, and can be applied only to low frequency coefficients in a specific region on the upper left side in order to maintain low complexity. The quadratic transformation applied to such a low frequency coefficient may be referred to as NSST (Non-Separable Section Condary Transfer), LFNST (low frequency non-separable transform), or RST (reduced separable trans). .. Further, the primary transformation may be referred to as a core transformation.

In one embodiment of the present invention, the primary conversion candidate used for the primary conversion and the secondary conversion kernel used for the secondary conversion may be predefined in various combinations. In the present specification, the primary conversion candidate used for the primary conversion may be referred to as an MTS candidate, but is not limited to the name. As an example, the primary conversion candidate may be a combination of conversion kernels (or conversion types) applied in each of the horizontal and vertical directions, and the conversion kernel may be any one of DCT2, DST7 and / or DCT8. It may be one. In other words, the primary conversion candidate can be at least one combination of DCT2, DST7 and / or DCT8. A specific example will be described below.

-Combination A

In combination A, as shown in Table 12 below, the primary conversion candidate and the secondary conversion kernel are defined by the intra prediction mode.

Referring to Table 12, as an example (Case1), when the intra prediction mode has directionality, two primary conversion candidates are used, and when there is no directionality (for example, DC, planner mode), four primarys are used. Conversion candidates are used. Here, the secondary conversion candidate can include two conversion kernels regardless of the direction of the intra prediction mode. That is, as described above, the intra-prediction mode predefines a plurality of secondary transformation kernel sets, and each of the predefined secondary transformation kernel sets includes two transformation kernels.

Further, as an example (Case2), when the intra prediction mode has directionality, two primary conversion candidates are used, and when there is no directionality, four primary conversion candidates are used. Here, the secondary conversion candidate can include one conversion kernel when the intra prediction mode has directionality, and may include two conversion kernels when it does not have directionality.

Further, as an example (Case3), when the intra prediction mode has directionality, two primary conversion candidates are used, and when there is no directionality, four primary conversion candidates are used. Here, the secondary conversion candidate can include one conversion kernel regardless of the direction of the intra prediction mode.

-Combination B

In combination B, the primary conversion candidate and the secondary conversion kernel are defined by the intra prediction mode as shown in Table 13 below.

Referring to Table 13, as an example (Case1), three primary conversion candidates are used regardless of the directionality of the intra prediction mode. Here, the secondary conversion candidate can include two conversion kernels regardless of the direction of the intra prediction mode. That is, as described above, the intra-prediction mode predefines a plurality of secondary transformation kernel sets, and the predefined plurality of secondary transformation kernel sets can each include two transformation kernels.

Further, as an example (Case2), three primary conversion candidates are used regardless of the direction of the intra prediction mode. Here, the secondary conversion candidate can include one conversion kernel when the intra prediction mode has directionality, and may include two conversion kernels when it does not have directionality.

Further, as an example (Case3), three primary conversion candidates are used regardless of the direction of the intra prediction mode. At this time, the secondary conversion candidate can include one conversion kernel regardless of the direction of the intra prediction mode.

-Combination C

In combination C, the primary conversion candidate and the secondary conversion kernel are defined by the intra prediction mode as shown in Table 14 below.

Referring to Table 14, as an example (Case1), when the intra prediction mode has directionality, two primary conversion candidates are used, and when there is no directionality (for example, DC, planner mode), three primarys are used. Conversion candidates are used. Here, the secondary conversion candidate can include two conversion kernels regardless of the direction of the intra prediction mode. That is, as described above, the intra-prediction mode predefines a plurality of secondary transformation kernel sets, and the predefined plurality of secondary transformation kernel sets can each include two transformation kernels.

Further, as an example (Case2), when the intra prediction mode has directionality, two primary conversion candidates are used, and when there is no directionality, three primary conversion candidates are used. Here, the secondary conversion candidate can include one conversion kernel when the intra prediction mode has directionality, and may include two conversion kernels when it does not have directionality.

Further, as an example (Case3), when the intra prediction mode has directionality, two primary conversion candidates are used, and when there is no directionality, three primary conversion candidates are used. Here, the secondary conversion candidate can include one conversion kernel regardless of the direction of the intra prediction mode.

In the above, the case of using a plurality of primary conversion candidates has been mainly described. In the following, a combination of a primary conversion and a secondary conversion for the case of using a fixed primary conversion candidate will be described as an example.

-Combination D

In the combination D, the primary conversion candidate and the secondary conversion kernel are defined by the intra prediction mode as shown in Table 15 below.

Referring to Table 15, as an embodiment, one primary conversion candidate is fixedly used regardless of the intra prediction mode. For example, the fixed primary conversion candidate can be at least one combination of DCT2, DST7 and / or DCT8.

As an example (Case1), one primary conversion candidate is fixedly used regardless of the intra prediction mode, where the secondary conversion candidate includes two conversion kernels regardless of the direction of the intra prediction mode. Can be done. That is, as described above, the intra-prediction mode predefines a plurality of secondary transformation kernel sets, and the predefined plurality of secondary transformation kernel sets can each include two transformation kernels.

Further, as an example (Case2), one primary conversion candidate is fixedly used regardless of the intra prediction mode, and here, the secondary conversion candidate uses one conversion kernel when the intra prediction mode has directionality. If included and not directional, two conversion kernels can be included.

Further, as an example (Case3), one primary conversion candidate is fixedly used regardless of the intra prediction mode, and here, the secondary conversion candidate includes one conversion kernel regardless of the direction of the intra prediction mode. Can be done.

-Combination E

In the combination E, the primary conversion candidate and the secondary conversion kernel are defined by the intra prediction mode as shown in Table 16 below.

Referring to Table 16, the secondary transformation is defined only when DCT2 is applied as the primary transformation. In other words, if the MTS is not applied (ie, if the DCT2 is applied as the primary transformation), then the secondary transformation is applicable. As described with reference to FIG. 10, in the present specification, the case where MTS is applied and the case where MTS is not applied will be described separately, but the present invention is not limited to such expressions. For example, whether or not to apply MTS is whether to use another conversion type (or conversion kernel) other than the predetermined specific conversion type (which may be referred to as a basic conversion type, a default conversion type, etc.). It can have the same meaning as whether or not. If MTS is applied, another conversion type other than the basic conversion type (eg, one or more combinations of multiple conversion types) is used for the conversion and MTS is not applied. If so, the basic conversion type may be used for the conversion. In one embodiment, the basic conversion type may be set (or defined) to DCT2.

As an example (Case1), if DCT2 is applied to the primary transformation, the secondary transformation can be applied, in which case the secondary transformation candidates will have two transformation kernels regardless of the direction of the intra prediction mode. include. That is, as described above, the intra-prediction mode predefines a plurality of secondary transformation kernel sets, and the predefined plurality of secondary transformation kernel sets can each include two transformation kernels.

Further, as an example (Case2), when DCT2 is applied to the primary transformation, the secondary transformation is applicable, and here, the secondary transformation candidate uses one conversion kernel when the intra prediction mode has directionality. If included and not directional, two conversion kernels can be included.

Further, as an example (Case3), when DCT2 is applied to the primary transformation, the secondary transformation is applicable, and here, the secondary transformation candidate includes one conversion kernel regardless of the direction of the intra prediction mode. be able to.

FIG. 26 is a flowchart illustrating a method of converting a video signal according to an embodiment to which the present invention is applied.

With reference to FIG. 26, the decoder will be mainly described for convenience of explanation, but the present invention is not limited thereto, and the conversion method for the video signal according to the present embodiment is substantially the same in the encoder. Applicable. The flowchart of FIG. 26 is performed by the decoding device 200 or the inverse conversion unit 230.

The decoding device 200 parses a first syntax element indicating a primary transformation kernel currently applied to a primary transformation kernel (S2601).

The decoding device 200 determines whether or not a secondary transformation can be applied to the current block based on the first syntax element (S2602).

If a secondary transformation can be applied to the current block, the decoding device 200 parses a second syntax element indicating a secondary transformation kernel (secondary transformation kernel) applied to the secondary transformation of the current block (S2603). ..

The decoding device 200 performs a quadratic inverse transformation on a specific region on the upper left side of the current block using the quadratic transformation kernel indicated by the second geometric element, so that the block is quadratic inverse transformed. (S2604).

The decoding device 200 induces the residual block of the current block by performing the primary inverse transformation on the secondary inverse transformed block using the primary transformation kernel indicated by the first geometric element. (S2605).

As described above, the S2602 step is performed by determining that a quadratic transformation can be applied to the current block if the first syntax element represents a predefined first transform kernel. Here, the first conversion kernel is defined in DCT2.

Further, as described above, the decoding device 200 is used for the secondary conversion of the current block in the secondary conversion kernel set (secondary transition kernel set) defined in advance based on the intra prediction mode of the current block. The next conversion kernel set can be determined. The second syntax element can then indicate a secondary transformation kernel applied to the secondary transformation of the current block within the determined secondary transformation kernel set.

Also, as mentioned above, each of the predefined secondary conversion kernel sets can include two conversion kernels.

In one embodiment of the present invention, an example of a syntax structure in which an MTS (Multiple Transform Set) is used will be described.

As an example, Table 17 below shows an example of the syntax structure of a sequence parameter set.

Referring to Table 17, the availability of the MTS according to the embodiment of the present invention can be signaled through the sequence parameter set syntax. Here, sps_mts_intra_enabled_flag indicates whether the MTS flag or MTS index is present in the sublevel syntax (syntax) of the intracoding unit (eg, resilient coding syntax, conversion unit syntax). And sps_mts_inter_enabled_flag indicates whether the MTS flag or MTS index is present in the sub-level syntax of the intercoding unit.

As another example, Table 18 below shows an example of the conversion unit syntax structure.

Referring to FIG. 18, cu_mts_flag indicates whether MTS is applied to the resilient sample of the Luma conversion block. For example, if cu_mts_flag = 0, MTS is not applied to the register dual sample of the Luma conversion block, and if cu_mts_flag = 1, MTS is applied to the register dual sample of the Luma conversion block.

As described above, in the present invention, the case where the MTS is applied and the case where the MTS is not applied will be described separately based on the MTS and the flag, but the present invention is not limited to such expressions. For example, the applicability of MTS depends on whether to use a specific conversion type (or conversion kernel) other than a specific predefined conversion type (which can be referred to as a basic conversion type, default conversion type, etc.). It can have the same meaning. If MTS is applied, other conversion types other than the basic conversion type (eg, one or a combination of multiple conversion types) will be used for the conversion and MTS will not be applied. If the basic conversion type is used for the conversion. In one embodiment, the basic conversion type can be set (or defined) to DCT2.

As an example, the MTS flag syntax that indicates whether MTS is applied to the current conversion block and the MTS index syntax that indicates the conversion type applied to the current block are individually from the encoder. It may also be transferred to a decoder, as another example: whether MTS applies to the current conversion block, and if MTS applies, a syntax that includes all of the conversion types that apply to the current block. (For example, the MTS index) can also be transferred from the encoder to the decoder. That is, in the latter embodiment, a syntax (or syntax) that indicates the type of conversion applied to the current conversion block (or unit) within a group (or conversion type set) of complete conversion types, including the basic conversion types described above. The element) can be transferred from the encoder to the decoder.

Therefore, despite that representation, the syntax (MTS index) that indicates the type of transformation applied to the current transformation block can include information about whether to apply the MTS. That is, in the latter embodiment, only the MTS index without the MTS flag may be signaled, and in this case, it can be interpreted that DCT2 is included in MTS, but in the present invention, the case where DCT2 is applied is used. , MTS can be described as not applicable, and nevertheless, the descriptive scope of MTS is not limited to its definition.

As another example, Table 19 below shows an example of the syntax structure of the resilient unit.

Referring to Table 19, the transform_skip_flag and / or mts_idx syntax (or syntax element) can be signaled through the resilient syntax. However, this is an example, and the present invention is not limited thereto. For example, the transform_skip_flag and / or mts_idx syntax (syntax) can also be signaled through the transform unit syntax.

In the following, we propose a method to improve the complexity by applying the primary transform only to the predefined regions. Complexity can be increased if a combination of various different transformations (or transformation kernels) such as MTS (eg, DCT2, DST7, DCT8, DST1, DCT5, etc.) is selectively applied to the primary transformation. .. In particular, the larger the size of the coding block (or conversion block), the greater the complexity that can be significantly increased by having to consider a variety of transformations.

Therefore, in the embodiments herein, for the sake of reduced complexity, instead of performing (or applying) the transformation on all areas, on certain conditions, in the predefined areas. Only suggests a way to perform the conversion.

As an example, based on the reduced transform (RT) method described in FIGS. 16 to 24 above, the encoder has a primary transform for an MxM size pixel block (luma block). ) To acquire the conversion block of MxM size, instead of acquiring the conversion block of RxR size. As an example, the RxR region can be the top-left RxR region in the current block (coding block, conversion block). The decoder can acquire the MxM size conversion block by performing the reverse primary conversion only in the region of the size of RxR (M> = R).

As a result, non-zero coefficients can exist only in the RxR region. As an example, in this case, the decoder can zero-out the values of the coefficients existing in the region other than the RxR region without performing the calculation. The encoder can perform forward conversion so that only the RxR region remains (so that the effective coefficient can exist only in the RxR region).

Also, the decoder should apply the primary transformation (ie, inverse transformation) only to the predefined regions determined by the size of the coding block (or transformation block) and / or the type of transformation (or transformation kernel). Can be done. Next, Table 20 shows the Reduced Adaptive Multiple Transform using predefined values of R (which can be referred to as Reduced factor, Reduced transform factor, etc.) depending on the size of the transformation (or the size of the transform block). (RAMT) is illustrated. In the present invention, Reduced Adaptive Multiple Transform (RAMT) indicating a reduced transformation that is adaptively determined according to a block size is referred to as Reduced MTS (Multiple Transform Selection), Reduced explicit multiple transform, Reduced primary transform, and the like. be able to.

With reference to Table 20, at least one reduced transformation can be defined, depending on the size of the transformation (or the size of the transformation block). In one embodiment, which of the reduced transformations exemplified in Table 20 is used is the transformation (or transformation kernel) applied to the current block (coding block or transformation block). Can be determined based on. In Table 20, it is assumed that three reduced conversions are used, but the present invention is not limited thereto, and one or more various numbers of reduced conversions are used depending on the size of the conversion. Can be predefined.

Further, in applying the reduced adaptive multiple transform described above in the embodiment of the present invention, the reduced transform factor (R) can be determined depending on the linear transformation. For example, when the linear transformation is DCT2, the computational complexity is relatively simple compared to other linear transformations (eg, a combination of DST7 and / or DCT8), so for small blocks, the reduced transformation is used. The decrease in coding performance can be minimized by using a relatively large R value. The following Table 21 illustrates a Reduced Adaptive Multiple Transform (RAMT) with a predefined R value based on the size of the transformation (or the size of the transformation block) and the transformation kernel.

Referring to Table 21, if the transformation applied to the primary transformation is DCT2 and other transformations (eg, a combination of DST7 and / or DCT8), different Reduced transform factors are used.

FIG. 27 is a diagram illustrating a method of encoding a video signal using a reduced transform as an embodiment to which the present invention is applied.

Referring to FIG. 27, the encoder first determines whether to apply the transformation to the current block (S2701). The encoder can encode the transform skip flag based on the determined result. In this case, the step of encoding the conversion skip flag can be included in the S2701 step.

When a transformation is applied to the current block, the encoder determines the transformation kernel to be applied to the primary transform of the current block (S2702). The encoder can encode a transform index that points to a determined transform kernel, in which case a step of encoding the transform index can be included in the S2702 step.

The encoder determines the region in which the effective coefficient exists in the current block based on the transformation kernel applied to the primary transformation of the current block and the size of the current block (S2703).

Also, as an embodiment, the encoder is a conversion in which the conversion kernel indicated by the conversion index is a predefined conversion, and the width and / or height of the current block is a predefined size. If it is larger, a region having the width and / or height of the predefined size can be determined as the region in which the effective factor is present.

For example, the predefined conversion may be any one of a plurality of conversion combinations configured with a combination of DST7 and / or DCT8, and the predefined size may be 16. Alternatively, the predefined conversion may be the remaining conversion excluding DCT2. Also, as an example, the encoder is a region where the width and / or height is 32 when the conversion kernel indicated by the conversion index is DCT2 and the width and / or height of the current block is greater than 32. Can be determined in the region to which the primary transformation is applied.

Further, as an embodiment, when the conversion kernel indicated by the conversion index belongs to the first conversion group, the encoder determines the smaller value of the current block width (width) and the first threshold value. The width of the area to which the linear transformation is applied is determined, and the height of the current block and the smaller value of the first threshold values are determined to be the height of the area where the effective coefficient exists. can do. As an example, the first threshold may be 32, but the present invention is not limited thereto, and may be 4, 8 or 16 as in Table 20 or Table 21 described above.

Then, when the conversion kernel specified by the conversion index belongs to the second conversion group, the encoder sets the smaller value of the current block width and the second threshold value in the region to which the primary conversion is applied. The width can be determined, and the smaller of the current block height and the second threshold can be determined as the height of the region where the effective coefficient exists. As an example, the second threshold can be 16, but the present invention is not limited thereto, and may be 4, 6, 8, 12, 32 as in Table 20 or Table 21 described above. There is.

As an embodiment, the first conversion group may include DCT2, and the second conversion group may include a combination of multiple conversions composed of a combination of DST7 and / or DCT8.

The encoder performs a forward primary transform using the transform kernel applied to the primary transform of the current block (S2704). The encoder can obtain the linearly transformed conversion coefficient in the region where the effective coefficient exists by performing the forward linear transformation. As an embodiment, the encoder can apply a secondary transform to the first-order transformed coefficients, in which case the methods described above with FIGS. 6-26 can be applied. ..

FIG. 28 is a diagram illustrating a method of decoding a video signal by using a reduced transform as an embodiment to which the present invention is applied.

The decoder checks whether the transform skip is applied to the current block (S2801).

The decoder acquires a transform in-dex indicating the transform kernel applied to the current block from the video signal if the transform skip is not applied to the current block (S2802).

The decoder has a primary transform to the current block based on the transform kernel indicated by the transform index and the size (ie, width and / or height) of the current block. The area to be applied is determined (S2803).

As an embodiment, the decoder can consider in the current block the coefficients of the remaining regions other than the region to which the linear transformation is applied as zero.

Also, as an embodiment, the decoder is a conversion in which the conversion kernel indicated by the conversion index is a predefined conversion, and the width and / or height of the current block is a predefined size. If it is larger, a region having the width and / or height of the predefined size can be determined as the region to which the primary transformation is applied.

For example, the predefined conversion may be any one of a plurality of conversion combinations configured with a combination of DST7 and / or DCT8, and the predefined size may be 16. Alternatively, the predefined conversion may be the remaining conversion excluding DCT2. Also, as an example, the decoder is a region where the width and / or height is 32 when the conversion kernel indicated by the conversion index is DCT2 and the width and / or height of the current block is greater than 32. Can be determined in the region to which the primary transformation is applied.

Further, as an embodiment, when the conversion kernel indicated by the conversion index belongs to the first conversion group, the decoder determines the smaller value of the current block width (width) and the first threshold value. The width of the area to which the primary transformation is applied is determined, and the height of the current block and the smaller value of the first threshold values are set to the height of the area to which the primary transformation is applied. Can be decided. As an example, the first threshold may be 32, but the present invention is not limited thereto, and may be 4, 8 or 16 as in Table 20 or Table 21 described above.

Then, when the conversion kernel specified by the conversion index belongs to the second conversion group, the decoder uses the smaller value of the current block width and the second threshold value as the area to which the primary conversion is applied. The width can be determined, and the smaller of the current block height and the second threshold can be determined to be the height of the area to which the linear transformation is applied. As an example, the second threshold may be 16, but the present invention is not limited thereto, and may be 4, 6, 8, 12, 32 as in Table 20 or Table 21 described above.

As an embodiment, the first conversion group may include DCT2, and the second conversion group may include a combination of multiple conversions composed of a combination of DST7 and / or DCT8.

The decoder performs an inverse primary transform on the region to which the primary transformation is applied, using the transformation kernel indicated by the transformation index (S2804). The decoder can obtain the conversion coefficient obtained by performing the first-order inverse transformation. As an embodiment, the decoder can apply a secondary transform to the inverse quantized transformation factor before performing the primary transformation, in which case it will be described in FIGS. 6-26 above. The method used can be applied.

According to the embodiments of the present specification, the worst case complexity can be significantly reduced by performing the transformation only on the predefined regions, depending on the specific conditions.

Further, in one embodiment of the present specification, when the MTS (EMT or AMT) flag is 0 (that is, when all DCT-2 conversions in the horizontal (horizontal) direction and the vertical (vertical) direction are applied). The encoder / decoder should perform (ie, consider or set) zero-out for high frequency components, leaving only 32 coefficients from the left and top, respectively, in the horizontal and vertical directions. Can be done. This example is referred to as a first embodiment for convenience of explanation in the examples described later, but the examples of the present specification are not limited thereto.

For example, in the case of a 64x64 TU (or CU), the encoder / decoder can leave the conversion coefficient only in the top-left 32x32 region and zero out the coefficients in the remaining region. Further, in the case of 64x16 TU, the encoder / decoder can leave the conversion coefficient only in the upper left end 32x16 region and perform zero-out to the coefficient in the remaining region. Further, in the case of 8x64 TU, the encoder / decoder can leave the conversion coefficient only in the upper left end 8x32 region and perform zero-out to the coefficient in the remaining region. That is, it is possible to set the conversion coefficient to exist only up to a maximum length of 32 in all the horizontal and vertical directions, whereby the conversion efficiency can be improved.

As an example, such a zero-out method may apply only the residual signal to which the intra-prediction is applied, or only the residual signal to which the inter-prediction is applied, so that the intra-prediction is applied. It can also be applied to all applied residual signals and residual signals to which interprediction is applied.

Further, in the embodiment of the present specification, when the MTS flag is 1 (that is, other conversions other than the horizontal and vertical DCT-2 conversions (for example, DST-7 or DCT-8) are applied). If possible), the encoder / decoder may perform (ie, consider or set) zero-out for the remaining high frequency components, leaving the coefficients in a particular region at the top left corner. can. This example is referred to as a second embodiment for convenience of explanation in the examples described later, but the examples of the present specification are not limited thereto.

As an embodiment, the encoder / decoder can be configured to leave only the conversion factor region of a portion of the upper left corner region, as in the following example. That is, the encoder / decoder can preset the length (or number) of horizontal and / or vertical transformation coefficients to which the primary transformation is applied, depending on the width and / or height. As an example, coefficients beyond the length to which the linear transformation is applied can be zeroed out.

--When the width (w) is the same as or larger than 2n, the conversion coefficient is left only for the length of w / 2p from the left, and the conversion coefficient of the remaining area is fixed (or set, regarded) to the value of 0 (zero). -out) can be made.

-When the height (h) is 2 m, the conversion coefficient can be left only for the h / 2q length from the top, and the remaining conversion coefficient can be fixed to a value of 0.

As an example, the values of m, n, p, q can be predefined to various values. For example, the values of m, n, p, q can be set to integer values equal to or greater than 0. Alternatively, it can be specifically set as in the following example.

1) (m, n, p, q) = (5, 5, 1, 1)

2) (m, n, p, q) = (4, 4, 1, 1)

For example, in the case of 1) defined in advance in the configuration, the conversion coefficient may remain only in the region of the upper left end 16x16 in the 32x16 TU, and the conversion coefficient remains only in the region of the upper left end 8x16 in the 8x32 TU. There is.

As an example, such a zero-out method may be applied only to the residual signal to which the intra-prediction is applied, or may be applied only to the residual signal to which the inter-prediction is applied, and may be applied to the intra-prediction. It can also be applied to all the residual signals to which the prediction is applied and the residual signals to which the inter-prediction is applied.

Further, in another embodiment of the present specification, when the MTS and the flag are 1 (that is, other conversions other than the horizontal and vertical DCT-2 conversions (for example, DST-7 or DCT-8)). (If applicable), the encoder / decoder may perform (ie, consider or set) zero-out for the remaining high frequency components, leaving the coefficients in a particular region at the top left corner. can. Or, more specifically, the encoder could zero-out the remaining high frequency components, leaving the coefficients in a specific region at the upper left corner predefined, and the decoder was zeroed out. Coding can be performed using the coefficients of the regions that have been recognized in advance and have not been zeroed out. However, the embodiment of the present specification is not limited to this, and the zero-out process on the decoder side can be understood as a process of regarding (and recognizing and setting) the zero-out region as zero. This example is referred to as a third embodiment for convenience of explanation in the examples described later, but the examples of the present specification are not limited thereto.

As an embodiment, the encoder / decoder can be configured to leave only the conversion factor region of a portion of the upper left corner region, as in the following example. That is, the encoder / decoder can preset the length (or number) of the horizontal and / or vertical conversion coefficients to which the primary transformation is applied, depending on the width and / or height. As an example, coefficients beyond the length to which the linear transformation is applied can be zeroed out.

-If the height (h) is equal to or greater than the width (w) and the height is equal to or greater than 2n, only the upper left corner wx (h / 2p) region is left with a conversion factor and the remaining region has a conversion factor of 0. Can be fixed (or set, considered) (zero-out) to the value of.

-When the width (w) is larger than the height (h) and the width is 2 m, the conversion coefficient is left only in the upper left end (w / 2q) xh region, and the remaining conversion coefficient is fixed to a value of 0. Can be done.

In the above example, when the height (h) and the width (w) are the same, the vertical direction is reduced (h / 2p), but the horizontal direction is set to be reduced (w / 2q). You can also do it.

As an example, the values of m, n, p, q can be predefined to various values. For example, the values of m, n, p, and q can be set to integer values that are the same as or greater than 0. Alternatively, it can be specifically set as in the following example.

(m, n, p, q) = (4, 4, 1, 1)

2) (m, n, p, q) = (5, 5, 1, 1)

For example, in the case of the configuration of No. 1), the conversion coefficient can remain only in the region of the upper left end 16x16 in the 32x16 TU, and the conversion coefficient remains only in the region of the upper left end 8x8 in the 8x16 TU. Sometimes.

As an example, such a zero-out method may be applied only to the residual signal to which the intra-prediction is applied, or may be applied only to the residual signal to which the inter-prediction is applied, and may be applied to the intra-prediction. It can also be applied to all of the residual signals to which the prediction is applied and the residual signals to which the inter-prediction is applied.

As described above, the first embodiment relating to the method of limiting the region of the conversion coefficient when the MTS and the flag are 0, and the second and third embodiments relating to the method of limiting the region of the conversion coefficient when the MTS and the flag are 1. Can be applied individually or in combination.

As an embodiment, the following combinations of configurations can be applied.

1st Example + 2nd Example

1st Example + 3rd Example

As previously described in the second and third embodiments, as an embodiment, such a zero-out method may be applied only to the residual signal to which the intra-prediction is applied, and the inter-prediction may be applied. It may be applied only to the applied residual signal, or it may be applied to all the residual signal to which the intra-prediction is applied and the residual signal to which the inter-prediction is applied. Therefore, when the MTS flag is 1, the following combination configuration can be applied. At this time, when the MTS flag is 0, the above-mentioned first embodiment can be applied.

In one embodiment of the specification, the encoder / decoder may not perform resilient coding on a region where the conversion factor is considered to have a value of 0 in response to the zero out in the previous embodiment. That is, the encoder / decoder can be defined to execute the resilient coding only in the area excluding the zero-out area.

In the first embodiment, the second embodiment, and the third embodiment described above, the region (or coefficient) having only a value of 0 in the TU is clearly determined. That is, the upper left corner region where the existence of the conversion coefficient is allowed is excluded, and the rest is zeroed out to the value of 0. Therefore, in the entropy coding (or register dual coding) process, the encoder / decoder is configured to bypass the region that is guaranteed to have a value of 0 without performing register dual coding. be able to.

In one embodiment, the encoder / decoder can code a flag (referred to as subblock_flag) (or syntax, syntax element) in the CG (Coefficient Group) indicating whether or not a non-zero conversion factor exists. Here, the CG can be set as a sub-block of the TU to a 4x4 or 2x2 block depending on the shape of the TU block and / or whether it is a color difference / luminance component.

At this time, the encoder / decoder can scan the inside of the CG and code the coefficient value (or the level value of the coefficient) only when the subblock_flag is 1. Therefore, for a CG that belongs to a region that is zeroed out to a value of 0, the encoder / decoder can be configured to have a value of 0 by default without subblock_flag coding.

Also, in one embodiment, the encoder may first code the position of the last located coefficient (or the syntax, syntax element indicating the position of the last valid coefficient) in the forward scan order. For example, the encoder can code the horizontal position last_coefficient_position_x and the vertical position last_coefficient_position_y.

The maximum value that last_coefficient_position_x and last_coefficient_position_y can have may be determined by the TU values (width -1) and (height -1), respectively, but there are non-zero coefficients due to zero out. If the area that can be limited is limited, the maximum value that last_coefficient_position_x and last_coefficient_position_y can have can also be limited.

Therefore, the encoder / decoder can code after limiting the maximum value that last_coefficient_position_x and last_coefficient_position_y can have in consideration of zero out. For example, the binarization method applied to last_coefficient_position_x and last_coefficient_position_y is the truncated unary (or truncated rice (TR), truncated binary (TB)) binarization method. In some cases, the encoder / decoder can adjust (reduce) the maximum length of the cut unary code to accommodate the adjusted (ie, can have last_coefficient_position_x and last_coefficient_position_y) maximum values.

Although some of the examples of the present invention described above have been described separately for convenience of explanation, the present invention is not limited thereto. That is, the embodiments described above may be executed independently, or a combination of one or more different examples may be executed.

FIG. 29 is a flowchart illustrating a method of decoding a video signal based on the reduced transform according to the embodiment of the present specification.

With reference to FIG. 29, the decoder will be mainly described for convenience of explanation, but the present invention is not limited thereto, and the method of converting a video signal according to the present embodiment is substantially the same for the encoder. Can be applied as such. The flowchart of FIG. 29 can be executed by the decoding device 200 or the inverse conversion unit 230.

The decoder obtains a transform index from the video signal indicating transform ker-nels applied in the horizontal and vertical directions of the current block (S2901).

The decoder determines the region to which the transformation (ie, inverse transformation) is applied to the current block based on the size of the transformation kernel and the current block indicated by the transformation index (S2902).

The decoder considers the coefficients of the remaining regions in the current block other than the region to which the transformation is applied as 0 (S2903).

The decoder performs an inverse transform on the region to which the transformation is applied using the transformation kernel indicated by the transformation index (S2904).

As mentioned above, the step of determining the region to which the transformation applies is the transformation in which the transformation kernel indicated by the transformation index is predefined, the width and / or height of the current block. If the height is greater than the predefined size, this can be done by determining a region having the width and / or height of the predefined size as the region to which the transformation applies.

As described above, the predefined conversion may be any one of a plurality of conversion combinations composed of combinations of DST7 and / or DCT8.

As mentioned above, the predefined size can be 16.

As mentioned above, the step of determining the region to which the transformation is applied is, if the transformation kernel indicated by the transformation index belongs to the first transformation group, of the current block width and 32. The smaller value is determined to be the width of the area to which the transformation is applied, and the smaller of the current block height (height) and 32 is determined to be the height of the area to which the transformation is applied. If the transformation kernel indicated by the transformation index belongs to a second transformation group, a smaller value of the width of the current block and 16 is determined to be the width of the area to which the transformation is applied and of the current block. Of the height and 16, the smaller value can be determined by determining the height of the region to which the transformation is applied. As an embodiment, the first conversion group may include DCT2, and the second conversion group may include a combination of multiple conversions composed of a combination of DST7 and / or DCT8.

As mentioned above, the coefficient to which the inverse transformation is applied further comprises the step of acquiring a syntax element indicating the position of the last significant coefficient in the scan order in the current block. Can be obtained from the video signal based on the position of the effective factor of.

As described above, the syntax element is binarized in a truncated unary manner, and the maximum value of the syntax element can be determined based on the region considered to be zero.

FIG. 30 shows an example of a block diagram of a device for processing a video signal as an embodiment to which the present invention is applied. The video signal processing device of FIG. 30 can correspond to the encoding device of FIG. 1 or the decoding device of FIG.

The video processing apparatus 3000 that processes a video signal includes a memory 3020 that stores the video signal and a processor 3010 that processes the video signal while being combined with the memory.

The processor 3010 according to the embodiment of the present invention can be configured by at least one processing circuit for processing a video signal, and by executing a command for encoding or decoding the video signal, the video signal can be processed. Can be processed. That is, the processor 3010 can encode the original video data or decode the encoded video signal by executing the encoding or decoding method described above.

In addition, the processing method to which the present invention applies can be produced in the form of a program executed by a computer and can be stored in a computer-readable storage medium. Multimedia data having the data structure according to the present invention can also be stored in a recording medium that can be read by a computer. The computer-readable recording medium includes all types of storage devices and distributed storage devices in which computer-readable data is stored. Computer-readable recording media include, for example, Blu-ray discs (BDs), Univers leaving serial buses (USB), ROMs, PROMs, EPROMs, EEPROMs, RAMs, CD-ROMs, magnetic tapes, floppy disks, and optical data. Storage equipment can be included. Also, the computer-readable recording medium includes media realized in the form of a carrier wave (eg, transmission over the Internet). In addition, the bitstream generated by the encoding method can be stored in a computer-readable recording medium or transferred via a wireless communication network.

Further, an embodiment of the present invention can be realized as a computer program product using a program code, and the program code can be executed on a computer according to the embodiment of the present invention. The program code can be stored on a computer-readable carrier.

As described above, the embodiments described in the present invention can be realized and implemented on a processor, microprocessor, controller, or chip. For example, the functional units shown in each figure can be realized and executed on a computer, processor, microprocessor, controller, or chip.

Further, the decoder and encoder to which the present invention is applied are real-time communication devices such as multimedia broadcast transmitter / receiver, mobile communication terminal, home cinema video device, digital cinema video device, surveillance camera, video dialogue device, video communication and the like. , Mobile streaming device, storage medium, camcorder, video on-demand (VoD) service provider, OTT video (Over the top video) device, Internet streaming service provider, 3D (3D) video device, videophone, video device, And can be included in medical video equipment and the like, and are used to process video or data signals. For example, an OTT video (Over the top video) device can include a game machine, a Blu-ray player, an Internet-connected TV, a home theater system, a smartphone, a tablet PC, a DVR (Digital Video Recoder), and the like.

In addition, the processing method to which the present invention applies can be produced in the form of a program executed by a computer and can be stored in a computer-readable storage medium. Multimedia data having the data structure according to the present invention can also be stored in a storage medium readable by a computer. The computer-readable recording medium includes all types of storage devices and distributed storage devices in which computer-readable data is stored. Computer-readable recording media include, for example, blue example discs (BDs), universal serial buses (USB), ROMs, PROMs, EPROMs, EEPROMs, RAMs, CD-ROMs, magnetic tapes, floppy disks, and optical data. A storage device can be included. Also, the computer-readable recording medium includes media realized in the form of a carrier wave (eg, transmission over the Internet). In addition, the bitstream generated by the encoding method can be stored in a computer-readable recording medium or transferred via a wireless communication network.

Further, an embodiment of the present invention can be realized as a computer program product using a program code, and the program code can be executed on a computer according to the embodiment of the present invention. The program code can be stored on a computer-readable carrier.

In the embodiment described above, the components and features of the present invention are combined in a predetermined form. Each component or feature should be considered selectively unless otherwise explicitly mentioned. Each component or feature can be implemented in a form that is not combined with other components or features. It is also possible to combine some components and / or features to form an embodiment of the invention. The order of operations described in the embodiments of the present invention can be changed. Some configurations or features of any embodiment may be included in other embodiments or may be substituted for configurations or features corresponding to other embodiments. It is self-evident that claims that have no explicit citation relationship can be combined to form an embodiment or can be included as a new claim by post-application amendment.

The embodiments according to the present invention can be realized by various means such as hardware, firmware, software or a combination thereof. In the case of hardware implementation, one embodiment of the present invention is one or more ASICs (application specific integrated circuits), DSPs (digital signal processors), DSPDs (digital signal processing devices), PLDs (programmable logic devices). , FPGAs (field programmable gate arrays), processors, controllers, microcontrollers, microprocessors and the like.

In the case of realization by firmware or software, one embodiment of the present invention can be realized in the form of a module, procedure, function or the like that executes the function or operation described above. The software code can be stored in memory and driven by the processor. The memory is located inside or outside the processor and can transmit and receive data to and from the processor by various means already known.

It is obvious to those skilled in the art that the invention can be embodied in other particular embodiments without departing from the essential features of the invention. Therefore, the detailed description given above should not be construed in a restrictive manner in all respects and should be regarded as exemplary. The scope of the invention should be determined by the reasonable interpretation of the appended claims, and all modifications within the equivalent scope of the invention are within the scope of the invention.

As described above, the preferred embodiment of the present invention described above has been disclosed for the purpose of exemplification, and those skilled in the art will be described below with the technical idea of the present invention disclosed in the appended claims. Within its technical scope, various other embodiments may be improved, modified, replaced or added.

Claims (14)

In the method of decoding a video signal based on a reduced transform,
The stage of acquiring a transform index from the video signal that indicates the transform kernels applied in the horizontal and vertical directions of the current block, and
The transformation is applied in the step of determining the region to which the transformation is applied to the current block and within the current block, based on the transformation kernel indicated by the transformation index and the size of the current block. The stage where the coefficients of the remaining areas other than the area are regarded as 0, and
A method of decoding a video signal, which comprises performing an inverse transform on a region to which the transformation is applied, using the transformation kernel indicated by the transformation index. The stage of determining the area to which the transformation is applied is
If the transformation kernel indicated by the transformation index is a predefined transformation and the width and / or height of the current block is greater than the predefined size, the said. The method of decoding a video signal according to claim 1, which is performed by determining a region having a width and / or height of a predefined size as a region to which the conversion is applied. The video signal according to claim 2, wherein the predefined conversion is any one of a plurality of conversion combinations composed of a combination of DST7 and / or DCT8. Method. The method of decoding a video signal according to claim 2, wherein the predefined size is 16. The stage of determining the area to which the transformation is applied is
If the conversion kernel specified by the conversion index belongs to the first conversion group, the smaller of the current block width (width) and 32 is determined to be the width of the area to which the conversion is applied. The smaller of the current block height and 32 is determined to be the height of the area to which the transformation is applied.
If the transformation kernel indicated by the transformation index belongs to the second transformation group, a smaller value of the width of the current block and 16 is determined to be the width of the region to which the transformation is applied, and the width of the current block. The method of decoding a video signal according to claim 1, wherein a small value among the height and 16 is determined to be the height of the region to which the conversion is applied. It further includes the step of acquiring a syntax element that indicates the position of the last significant coefficient in the scan order within the current block.
The method of decoding a video signal according to claim 1, wherein the coefficient to which the inverse transformation is applied is obtained from the video signal based on the position of the last effective coefficient. The syntactic element is binarized by a truncated unary method and is binarized.
The method of decoding a video signal according to claim 6, wherein the maximum value of the syntax element is determined based on the region regarded as 0. In a device that decodes a video signal based on a reduced transform.
A memory for storing the video signal and
Including the processor combined with the memory
The processor
Obtaining a transform index from the video signal indicating the transform kernels applied in the horizontal and vertical directions of the current block,
Based on the conversion kernel indicated by the conversion index and the size of the current block, the area to which the conversion applies to the current block is determined.
In the current block, the coefficients of the remaining regions other than the region to which the transformation is applied are regarded as 0, and the coefficients are regarded as 0.
A video signal decoding device that performs an inverse transform on the region to which the transformation is applied using the transformation kernel indicated by the transformation index. The processor
If the transformation kernel indicated by the transformation index is a predefined transformation and the width and / or height of the current block is greater than the predefined size, then the predefined. The video signal decoding device according to claim 8, wherein a region having a width and / or height of a size is determined as a region to which the conversion is applied. The video signal decoding device according to claim 9, wherein the predefined conversion is any one of a plurality of conversion combinations composed of a combination of DST7 and / or DCT8. .. The video signal decoding device according to claim 9, wherein the predefined size is 16. The processor
If the conversion kernel specified by the conversion index belongs to the first conversion group, the smaller of the current block width (width) and 32 determines the width of the area to which the conversion is applied. The height of the current block and the smaller of the 32 values are determined to be the height of the area to which the transformation is applied.
If the transformation kernel indicated by the transformation index belongs to the second transformation group, a smaller value of the width of the current block and 16 is determined to be the width of the region to which the transformation is applied, and the width of the current block. The video signal decoding device according to claim 8, wherein the smaller value of the height and 16 is determined to be the height of the region to which the conversion is applied. The processor
Acquires a syntax element that indicates the position of the last significant coefficient in the scan order within the current block.
The video signal decoding device according to claim 8, wherein the coefficient to which the inverse transformation is applied is obtained from the video signal based on the position of the last effective coefficient. The syntactic element is binarized by a truncated unary method and is binarized.
The video signal decoding device according to claim 13, wherein the maximum value of the syntax element is determined based on the region regarded as 0.

Images